Ant Colony Optimization (ACO) for Wireless Sensor. Networks ... protocol is simulated and tested using NS-2.35 simulator. Simulation based ...... Adaptive routing mainly defines the ability of the system, in calculating the optimized routes from ...
Computer Science
Anand Nayyar
Improvised Energy Efficient Routing Protocol based on Ant Colony Optimization (ACO) for Wireless Sensor Networks
Doctoral Thesis / Dissertation
Bibliographic information published by the German National Library: The German National Library lists this publication in the National Bibliography; detailed bibliographic data are available on the Internet at http://dnb.dnb.de .
This book is copyright material and must not be copied, reproduced, transferred, distributed, leased, licensed or publicly performed or used in any way except as specifically permitted in writing by the publishers, as allowed under the terms and conditions under which it was purchased or as strictly permitted by applicable copyright law. Any unauthorized distribution or use of this text may be a direct infringement of the author s and publisher s rights and those responsible may be liable in law accordingly.
Imprint: Copyright © 2017 GRIN Verlag, Open Publishing GmbH ISBN: 9783668599215
This book at GRIN: https://www.grin.com/document/385058
Anand Nayyar
Improvised Energy Efficient Routing Protocol based on Ant Colony Optimization (ACO) for Wireless Sensor Networks
GRIN Publishing
GRIN - Your knowledge has value Since its foundation in 1998, GRIN has specialized in publishing academic texts by students, college teachers and other academics as e-book and printed book. The website www.grin.com is an ideal platform for presenting term papers, final papers, scientific essays, dissertations and specialist books.
Visit us on the internet: http://www.grin.com/ http://www.facebook.com/grincom http://www.twitter.com/grin_com
IMPROVISED ENERGY EFFICIENT ROUTING PROTOCOL BASED ON ANT COLONY OPTIMIZATION (ACO) FOR WIRELESS SENSOR NETWORKS A THESIS submitted to DESH BHAGAT UNIVERSITY, MANDI GOBINDGARH In partial fulfillment of the requirement for the Award of the Degree of DOCTOR OF PHILOSOPHY in the field of
COMPUTER SCIENCE
DEPARTMENT OF COMPUTER SCIENCE
DESH BHAGAT UNIVERSITY, MANDI GOBINDGARH PUNJAB 2017
ACKNOWLEDGEMENT
It is my honour to express my deep gratitude to people who made this challenge possible for me. I would like to thank my Supervisor Dr. Rajeshwar Singh for giving me the opportunity to work under him. He offered me such a challenging topic for my thesis and with his strong support and constant guidance; I am finally able to achieve my goals. He did not only guide me in the course of this thesis, but also provided me an opportunity to be benefited from his vast knowledge. His enormous knowledge about wireless routing, was a source of immense help and it is his time and guidance that made this thesis better. Secondly, I am thankful to Dr. Sidhu Mam, Incharge of Research & Development Cell, Desh Bhagat University to provide me great support and helped me to overcome many difficulties in my research work. I am indebted to my parents, Mr. Sunil Kumar Nayyar and Mrs. Meenakshi Nayyar, my sister Dr. Sanchi Sharma and my best friend Er. Vikram Puri, for everything I own now. It is their prayers and support that gave me the courage and ability to fulfill my goals.
(Anand Nayyar)
iii
ABSTRACT
Routing and Energy Efficiency is regarded as highly challenging area of Sensor networks. Significant advancements in Wireless Sensor Networks (WSNs) opens doors for wide implementation in real-time applications like Industrial Monitoring, Smart Cities development, Underwater monitoring operations, tracking objects and many more. Energy Efficient routing is regarded as the most challenging task. Sensor networks mostly operate in complex and dynamic environments and routing becomes tedious task to maintain as the network size increases. Lots of routing protocolsReactive, Proactive and Hybrid are proposed by researchers but every protocol faces some limitations in terms of energy, routing, packet delivery ratio and security. Therefore, to overcome all the routing issues, the trend has shifted to Biological based Algorithms like Swarm Intelligence based techniques. Ant Colony Optimization based routing protocols have demonstrated exceptional results in terms of performance when applied to WSN routing. This thesis outlines routing protocols in sensor networks, highlight the concept of Swarm Intelligence and presents various Ant Colony Optimization based routing protocols for sensor networks. In addition to this, we present Ant Colony based Energy Efficient routing protocol (IEEMARP = Improvised Energy Efficient Multipath Ant Based Routing Protocol) for sensor networks. The proposed protocol takes into consideration various performance metrics like Packet Delivery Ratio, Throughput, Energy Efficiency, Routing Overhead and End-to-End delay. Proposed protocol is simulated and tested using NS-2.35 simulator. Simulation based results stated that IEEMARP routing protocol is overall 16% more efficient in terms of Packet delivery ratio, Energy Efficiency, Throughput, Routing Overhead and End-toEnd delay as compared to other ACO based routing protocols. In addition to this, IEEMARP
is
highly
reliable
protocol
to
ensure
timely
delivery
with
acknowledgement packet exchange between source node to sink node and vice versa and also combats the issue of congestion and packet dropping to large extent.
iv
TABLE OF CONTENTS
Acknowledgments....................................................................................................... iii Abstract ....................................................................................................................... iv Table of Contents ........................................................................................................ v List of Figures .............................................................................................................. x List of Tables............................................................................................................. xiii Acronyms ................................................................................................................... xv Page No. Chapter – 1 : Introduction
1-55
1.1
Sensor Networks- Evolution and Introduction
1
1.2
Model of Wireless Sensor Network
4
1.3
Wireless Sensor Networks- Design Principles and Challenges
5
1.4
Wireless Sensor Networks – Types
7
1.5
Wireless Sensor Networks- Classifications
11
1.6
Wireless Sensor Network Architecture: Protocol Stack
12
1.7
Routing in Wireless Sensor Networks
14
1.7.1
15
Challenges connected to Routing in Wireless Sensor Networks
1.7.2
Classifications of Routing Algorithms/Protocols
16
1.8
Swarm and Swarm Intelligence (SI)
19
1.9
Ant Colony Optimization (ACO)
24
1.9.1
Introduction
24
1.9.2
Ants in Nature
25
1.9.3
Ants Stigmergic behavior
25
1.9.4
Real Ants v/s. Artificial Ants
26
1.9.5
Ant Colony Optimization Metaheuristic
28
v
Page No. 1.9.6
Mathematical Model of Ant Colony Optimization
29
1.9.7
Components of Ant Colony Optimization (ACO)
30
1.9.8
Ant Colony Optimization Algorithms
31
1.9.8.1 Ant System Algorithm
31
1.9.8.2 Ant Colony System (ACS)
32
1.9.8.3 MAX-MIN Ant System
34
1.9.8.4 Ant Lion Optimizer
34
Ant Colony Optimization- Working and Algorithm
35
1.9.9
1.10 Suitability of Ant Colony Optimization (ACO) based approach for
36
Developing Energy Efficient Routing Protocols for Wireless Sensor Networks 1.11
ACO Based Routing Protocols for Wireless Sensor Networks
38
1.11.1 Sensor Driven Cost-Aware Ant Routing (SC)
38
1.11.2 Energy Efficient Ant Based Routing (EEABR)
39
1.11.3 Flooded Forward Ant Routing (FF)
40
1.11.4 Flooded Piggyback Ant Routing (FP)
41
1.11.5 Energy-Delay Ant Based (E-D Ants)
42
1.11.6 Ant Colony Based Reinforcement Learning Algorithm
43
(AR and IAR) 1.11.7 Basic Ant Based Routing (BABR) for Wireless Sensor
45
Networks (WSN) 1.11.8 Ant Based Quality of Service Routing (ACO-QoSR)
45
1.11.9 Ant Colony Optimization based Location-aware Routing
47
(ACLR) 1.12
Energy Efficient Routing Protocols based on Ant Colony
48
Optimization for Wireless Sensor Networks 1.12.1 Ant Chain Protocol
48
1.12.2 Ant Aggregation
48
1.12.3 Pheromone Based Energy Aware Directed Diffusion
49
(PEADD) vi
Page No. 1.12.4
Ant Colony Multicast Trees (ACMT)
49
1.12.5
Improvised Ant Colony Routing (IACR)
50
1.12.6
ACO Router Chip
50
1.12.7
Energy Balanced Ant Based Routing Protocol (EBAB)
51
1.12.8
Adaptive Clustering for Energy Efficient WSN based on
52
ACO (ACO-C) 1.12.9
Ant Colony Clustering Algorithm (ACALEACH)
1.12.10 Ant
Colony
Optimization
based-
Energy-Aware
52 52
Multipath Routing Algorithm (ACO-EAMRA)
1.13
1.12.11 Energy Efficient ACO Based QoS Routing (EAQR)
53
1.12.12 Comprehensive Routing Protocol (CRP)
53
Organization of Thesis
54
Summary
54
Chapter – 2 : Literature Review
56-75
Chapter – 3 : Research Methodology
76-84
3.1
Motivation
76
3.2
Research Problem
76
3.3
Research Objectives
79
3.4
Research Methodology
80
3.5
Research Contributions
81
3.6
Scope of Research
82
3.7
Research Gaps Identified
83
Chapter – 4 :
IEEMARP: Improvised Energy Efficient Multipath
85-97
Ant Colony Optimization based Routing Protocol for Sensor Networks 4.1
Problem Definition and Background
vii
85
Page No. 4.2
Protocol Design Choices
87
4.2.1
Energy Efficiency
87
4.2.2
Reliability
88
4.2.3
Dynamic Network and Scalability
88
4.2.4
Throughput and Routing Overhead
88
4.3
Assumptions
89
4.4
IEEMARP Protocol- Operation
90
4.4.1
Neighborhood Discovery via Link Knowledge
90
4.4.2
Forwarding of Packets / Fault Localization
90
4.4.3
Reliable End-to-End Communication from Source to Destination
91
4.5
IEEMARP Routing Protocol- Properties
93
4.6
IEEMARP Protocol- Algorithm
94
4.7
IEEMARP Routing Protocol- Algorithm
96
Summary Chapter – 5 : Simulation and Performance Analysis of IEEMARP
98-131
Routing Protocol 5.1
Introduction to NS-2 Simulator
98
5.2
Performance Metrics
99
5.3
Simulation and Performance Comparison of Basic Ant Colony
101
Optimization (ACO) Routing Protocol with AODV, DSR and DSDV Routing Protocols for Wireless Sensor Networks
5.4
5.3.1
Flowchart of Simple Ant Net Based Routing
101
5.3.2
Simulation Parameters
102
5.3.3
Simulation Scenarios
102
5.3.4
Simulation Results
106
Simulation and Performance Comparison of Basic Ant Colony Optimization based Routing Protocols: ACEAMR, AntChain, EMCBR and IACR viii
111
Page No.
5.5
5.4.1
Simulation Parameters
111
5.4.2
Simulation Results
112
5.4.3
Overall Analysis and Best Protocol Suitability
118
Simulation and Performance Comparison of Proposed Routing
118
Protocol 5.5.1
Simulation Parameters
118
5.5.2
Simulation Scenarios and IEEMARP Routing Protocol
119
working 5.5.3
Performance Results of IEEMARP Routing Protocol with
123
ACEAMR, AntChain, EMCBR and IACR routing protocols on performance metrics 5.5.4
Performance Comparison of IEEMARP routing protocol
130
with Traditional WSN routing protocols: DSR, DSDV and Basic ACO Summary
130
Conclusion and Future Scope
132-133
References
134-150
ix
LIST OF FIGURES
Figure No.
Description
Page No.
1.1
Wireless Sensor Network
5
1.2
Sensor Node components
6
1.3
Types of Wireless Sensor Networks
8
1.4
Terrestrial WSN
8
1.5
Underwater WSN
9
1.6
Underground WSN
9
1.7
Mobile WSN
10
1.8
Multimedia WSN
10
1.9
Protocol Stack of Wireless Sensor Networks
13
1.10
Classifications of Routing Algorithms
16
1.11
Ants stigmergic behavior
26
1.12
Ant Colony Optimization
30
3.1
Research Methodology Undertaken
80
4.1
IEEMARP Routing Protocol Operation
95
5.1
Basic Structure of NS-2 Simulation
99
5.2
Flowchart demonstrating Ant Based Routing Algorithm
101
5.3
Deployment of 100 Nodes creating Wireless Sensor Network Scenario
102
5.4
Base Station location
103
5.5
Route Discovery
103
5.6
Transmission of TCP Packet
104
5.7
Generation of Acknowledgement Packet
104
x
Figure No.
Description
Page No.
5.8
Control Packet Overhead
105
5.9
Handling of Route Failure by Ant Colony Optimization Algorithm
105
5.10
Packet Delivery Ratio comparing ACO with DSR, DSDV and AODV routing protocols.
107
5.11
Throughput comparing ACO with DSR, DSDV and AODV routing protocols
108
5.12
Routing Overhead reduced by ACO as compared to DSR, DSDV and AODV routing protocols
109
5.13
End to End significantly reduced by ACO as compared to DSR, DSDV and AODV routing protocols
110
5.14
Energy Consumption by Node utilized via ACO protocol as compared to DSR, DSDV and AODV routing protocols
111
5.15
Packet Delivery Ratio of ACEAMR, AntChain, EMCBR and IACR
113
5.16
Comparison of Routing Throughput parameters
114
5.17
Comparison of Routing Protocols Probability Routing Overhead
of
115
5.18
Energy Consumption comparison of Sensor Nodes on basis of different routing protocols in sensor networks
116
5.19
Comparison of Routing Protocols on basis of End-toEnd Delay
117
5.20
IEEMARP Simulation Start Scenario with Neighbouring Nodes Routing Table update using ACO routing protocol.
120
5.21
Route Discovery by IEEMARP Routing Protocol
120
xi
Protocols
considering
on
basis
Figure No.
Description
Page No.
5.22
Packet Transmission by Sensor Node to Sink Node via TCP Protocol for reliable end-to-end communication.
121
5.23
ACK Packet transmitted by Base Station (Node 0) to Sensor node confirming the acknowledgement of the received packet
122
5.24
Packet Dropped due to Routing Overhead
122
5.25
Link Status between Source Node to Sink Node.
123
5.26
Packet Delivery Ratio- Performance Metric Comparison
124
5.27
IEEMARP throughput comparison with other routing protocols
125
5.28
Routing Overhead based Performance comparison of IEEMARP routing protocol with other routing protocols.
126
5.29
Energy Consumption comparison of Routing Protocols with IEEMARP routing protocol
128
5.30
End-to-End delay based performance comparison of IEEMARP routing protocol with WSN based other routing protocols
129
xii
LIST OF TABLES
Table No.
Description
Page No.
1.1
Sensor Node Evolution
2
1.2
Real Ants V/s Artificial Ants.
27
5.1
Outlines the Simulation Parameters- Basic ACO Routing Protocol
102
5.2
Packet Delivery Ratio of ACO v/s DSDV, AODV and DSR Routing Protocols
106
5.3
Throughput analysis of ACO v/s AODV, DSDV and DSR routing protocols
107
5.4
Routing Overhead based results and comparison of ACO with DSDV, AODV and DSR routing Protocols.
108
5.5
End to End delay based results and comparison of ACO with DSDV, AODV and DSR routing Protocols.
109
5.6
Energy Consumption by Nodes in Path Selection and Data transmission between sender and receiver node.
110
5.7
Simulation Parameters for Performance Comparison of ACEAMR, IACR, EMCBR and AntChain Routing Protocol.
112
5.8
Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Packet Delivery Ratio
113
5.9
Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Throughput
114
5.10
Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Routing Overhead
115
xiii
Table No.
Description
Page No.
5.11
Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Energy Consumption
116
5.12
Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of End-to-End Delay
117
5.13
Overall Analysis of Best Protocol Suitability among ACEAMRA, EMCBR, IACR and AntChain Routing Protocol
118
5.14
Simulation Parameters for Simulating IEEMARP Routing Protocol and Comparison with ACEAMRA, AntChain, IACR and EMCBR routing protocols
119
5.15
Performance comparison of Routing ProtocolsACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Packet Delivery Ratio
124
5.16
Performance comparison of Routing ProtocolsACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Throughput
125
5.17
Performance comparison of Routing ProtocolsACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Routing Overhead
126
5.18
Performance comparison of Routing ProtocolsACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Energy Consumption
127
5.19
Performance comparison of Routing ProtocolsACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of End-to-End Delay
129
5.20
Performance Comparison of IEEMARP routing protocol with DSDV, DSR and Basic ACO routing protocols on varied performance parameters
130
xiv
ACRONYMS
ABC
Artificial Bee Colony
ABEBR
ACO based Energy balance routing algorithm
ACALEACH
Ant Colony Clustering Algorithm
ACEAMRA
Ant Colony Optimization based Energy Aware Multipath Routing Algorithm
ACK
Acknowledgement
ACLR
Ant Colony Optimization based Location-aware Routing
ACMT
Ant Colony Multicast Tress
ACO
Ant Colony Optimization
ACO-C
Adaptive Clustering for Energy Efficient WSN based on ACO
ACO-QoSR
Ant Based Quality of Service Routing
ACS
Ant Colony System
AI
Artificial Intelligence
AODV
Ad Hoc On demand Distance Vector
AR
Adaptive Routing
AS
Ant System
BANT
Backward Ant
CRP
Comprehensive Routing Protocol
DARPA
Defense Advanced Projects Agency
DSN
Distributed Sensor Networks
EAQR
Energy Efficient ACO Based QoS Routing
EBAB
Energy Balanced Ant Based Routing Protocol
E-D
Energy-Delay Ant
EEABR
Energy Efficient Ant Based Routing Algorithm xv
EMCBR
Efficient Minimum-Cost Bandwidth-Constraint Routing
EPWSN
Energy and Path aware Ant Colony Optimization based routing protocol
FANT
Forward Ant
FF
Flooded Forward
FP
Flooded Piggybacked
FS
Fish Swam
GPS
Global Positioning System
IACR
Improved Ant Colony Routing
IAR
Improved Adaptive Routing
IEEABR
Improvised Energy Efficient ant based routing
IEEMARP
Improvised Energy Efficient Multipath Ant based Routing Protocol
MANETS
Mobile Ad Hoc Networks
MCBCR
Minimum-Cost Bandwidth-Constrained Routing
MEMS
Microelectronic mechanical systems
MLBCR
Maximum Lifetime Bandwidth Constrained Routing
MMAS
MAX-MIN Ant System
MMBCR
Min-Max Battery Cost Routing
MTPR
Minimum Total Transmission Power Routing
NAM
Network Animator
NS-2
Network Simulator-2
PEADD
Pheromone Based Energy Aware Directed Diffusion
PSO
Particle Swarm Optimization
QoS
Quality of Service
RF
Radio Frequency xvi
SC
Sensor driven Cost-aware Ant Routing
SI
Swarm Intelligence
SOSUS
Sound Surveillance System
TCP
Transmission Control Protocol
TSP
Travelling salesman problem
WSN
Wireless Sensor Networks
xvii
Chapter – 1
Introduction This chapter outlines the concept of Wireless Sensor Networks- Evolution, Types, Design principles and challenges, classifications and protocol stack in addition to various Adaptive and Nonadaptive routing protocols for sensor networks and also highlights the comparison among protocols on varied performance parameters. The chapter also covers comprehensive details regarding Swarm Intelligence, Ant Colony Optimization and also enlists various ACO based Energy Efficient protocols for performing routing activity in sensor networks.
Wireless Sensor Networks Wireless Sensor Networks comprise of few to thousands of sensor nodes with sophisticated capabilities to interact with environment by sensing or controlling physical parameter with power of collaborating among each other to perform the tasks [1]. According to SmartDust program of DARPA, Wireless Sensor Network is [1]: “A Sensor Network is a deployment of massive numbers of small, inexpensive, selfpowered devices that can sense, compute and communicate with other devices for the purpose of gathering local information to make global decisions about a physical environment”.
1.1
Sensor Networks- Evolution and Introduction
The initial deployment of Sensor Network was done during Cold War by United States of America [4] when a large network comprising of acoustic sensors were deployed at strategic locations beneath the ocean to track Soviet Union submarines. The system of Acoustic sensors was defined as Sound Surveillance System (SOSUS). The sensor network deployed was not wireless all the sensors are connected via wired links and doesn‟t have any sort of energy constraints. Serious research on Sensor networks was started in late 1980s by Defense Advanced Projects Agency (DARPA) via Distributed Sensor Networks (DSN) program. DSN 1
program comprised of research surrounding acoustic sensors communication, advanced techniques regarding processing, algorithms and distributed software. Table 1.1. Sensor Node Evolution Parameter
1980-1990s
2000-2009
2010-Onwards
Companies
Custom Contractors
Crossbow, Sensoria
Libelium, Dust Inc, National Instruments, Texas Instruments
Size
Large shoe box
Pack of cards to small shoe box
Dust Particles
Node Architecture
Different sensing, processing and communication
Integrated sensing, processing and communication
High Speed, Reliable and Integrated sensing, processing and communication
Topology
Point-to-Point, Star
Client-Server, Peerto-Peer
Peer-To-Peer
Energy Source
Large batteries
AA Batteries
AAA Batteries, Solar Technology
With the advancements in area of Wireless communications, Digital Electronics and especially MEMS (Micro Electromechanical Systems), the development of smart sensors in terms of low cost, less energy consumption with multifunctional sensing has become possible. These smart sensors are extremely small size and have the capability of sensing, data processing and communicating with different sensor networks via Radio Frequency (RF) channel. Wireless Sensor Networks [2, 3] comprise of smart sensor nodes which are low power devices equipped with multifunctional sensors, processor, memory, power supply unit, a radio and an actuator. Different sorts of mechanical, thermal, environmental, magnetic, chemical and optical sensors are connected to sensor nodes to sense the data from real-time environment. Wireless sensor networks, in recent times, have seen great potential for many application areas like Military, Environmental monitoring, Medical, Space technology, Robotics, Industrial production and many more. [3,4,5]. 2
Wireless Sensor Networks have very less or no infrastructure [4]. It comprises of limited to tons of sensor nodes working cooperatively for monitoring a particular region and obtaining the live data from the environment. WSN‟s are of two types: Structured and Unstructured. In Structured WSN, the entire sensor nodes are deployed after proper planning. In Unstructured WSN, the sensor nodes are deployed in ad hoc manner in the environment, and after deployment, sensor nodes operate autonomously to perform the task of monitoring, sensing and reporting data. Wireless Sensor Networks, when compared to traditional networks, have its own limitations and constraints in terms of design and resources like limited energy, limited amount of energy, less processing capability, small communication range, less QoS and data storage capability. Research in WSNs is basically done in order to cope up with these challenges and to bring up sophisticated and advanced sensor nodes with new design via improvised routing protocols, new applications deployment and new algorithms to make sensor nodes less vulnerable to different types of attacks. The following are the basic features of sensor networks [7, 8]: a)
Capability for self-organizing and multi-hop routing.
b)
Dense deployment and intelligent cooperating communication capabilities with neighboring nodes.
c)
Dynamic nature of topology change in case of node failures or fading.
d)
Short range broadcast communication via radio frequency (RF).
e)
Limitations in terms of memory, computations and transmission power.
The vast growth of wireless communication not only alleviates the dependency on traditional wired networks, but also increases the suitability of mobile communication and enhances its computing power. In WSN, each sensor node acts as a router with ability of sensing. Although, the sensor nodes are highly mobile operating in dynamic changing topology, the protocol handling all the operation should be able to handle the rapid topology changes [9]. The sensor nodes are equipped with data processing communication capabilities. Sensor nodes collect the data via sensors integrated and transmits the data back to sink 3
node (Base Node) via Radio frequency (RF) directly or via Gateway node. Sensor Networks are usually small and inexpensive, so tons of nodes are deployed in random manner considering various limitations of energy, memory, computational speed and bandwidth. Every sensor utilizes tons of energy while receiving and transmitting data [10]. Extensive research is being undertaken by researchers to develop and propose energy efficient protocols cum algorithms for WSN to enhance the overall lifetime of network. In most of applications, sensor nodes are really constrained in terms of energy and bandwidth. Therefore, novel techniques to effectively utilize power and enhance bandwidth is required. Such constraints in WSN poses lots of challenges for deployment in real world applications [5] [6].
1.2
Model of Wireless Sensor Network
Like traditional ad hoc networks, WSN networks are limited in terms of resources, excessively prone to failures, so tons of sensor nodes are deployed in real time environment as compared to ad hoc networks. The topology of sensor networks is highly dynamic and changes continuously. The following are the major components of sensor networks:
Sensor deployment field: A sensor deployment field comprise of area in which nodes are deployed for real time sensing.
Sensor nodes: Sensor nodes perform the task of sensing, collecting, processing and routing the data back to sink node via path nodes.
Sink node: A sink node (also known as Base Node) perform the task of receiving, processing and storing the data received from deployed sensor nodes. They perform the task of reducing the total number of messages to be sent to maintain the energy level of nodes and overall network lifetime. Sink nodes are also termed as Data Aggregation nodes.
4
Fig. 1.1. Wireless Sensor Network
1.3
Wireless Sensor Networks- Design Principles and Challenges
It is clear from the above section that WSNs are application-specific, but even with wider applications deployment scope and technical advantages, there are some challenges of wireless sensor networks in effective design.
Effective topology control: Most of applications where sensor nodes are deployed remain static in the monitoring area and this is an important factor which distinguishes sensor networks from other ad hoc networks. Sensor networks require effective topology management when deployed to efficiently monitor the area. With issues of wireless interference, power and node damage, topology control seems to be indispensable for sensor nodes to operate in dynamic environment.
Sensor lifetime: Sensor nodes have limited energy and cannot remain operational for large period of time. The chief source providing power to sensor nodes is either AA or AAA battery cells and sometimes solar power. Most of the applications require WSNs to remain operational for longer period. Batteries with rechargeable capabilities can be a great support but requires consistent attention and continuous replacement which can boost up the cost of sensor 5
networks operations. So, energy efficiency improvement to enhance lifespan of WSNs has become the primary challenge for researchers across the world to overcome and even energy efficient routing protocols are proposed for improving sensor networks lifetime.
Cost: In order to make more real-time applications operational with sensor technology, the cost is the chief concern. The price of each sensor should be less than a dollar or ($1). The other components which are required for sensor node: power unit, sensing unit, processor and transceiver can even increase the cost.
Fig. 1.2. Sensor Node components
Fault Tolerance: Sensor nodes are prone to failure via hardware or software crash or sometimes via battery drain. Sensor network should be deployed with additional nodes and effective topology control strategy to make WSN network operate successfully regardless of node failures.
Standardization: Another issue directing the eyes of researchers for solutions in sensor networks is standardization. The design goal of WSNs is [6]: “The Goal of WSN engineers is to develop a cost-effective standards-based wireless networking solution that supports low-to-medium data rates, has low power consumption and guarantees security and reliability.” Varied standards are implemented for WSN like IEEE 802.15.4/ZigBee for indoor conditions and 6
IEEE 802.16/WiMAX for external environments. Standardizations is highly required in routing protocols of WSN as well as other areas of WSN like crosslayer, middleware‟s, operating systems, protocol stack etc.
MAC Layer Issues: MAC layer has direct impact on power utilization in terms of collision, packet overheead and idle listening. Power saving forward error control technique is very hard to implement due to high computing requirements, so in turn long data packets transmission is not possible in WSN.
Quality of Service (QoS): As WSN network works for real time and critical applications, it is hard to maintain QoS due to rapidly changing topology and routing information state which is consistently changing.
Security: Security is the key challenge issue in WSN. In sensor networks, it is utmost necessary for each sensor node and the base station to have the ability to verify that the data received is actually sent by trusted sender not any outside untrusted party. A false data can hamper the entire quality functioning of the network. Different sorts of threats in sensor networks are spoofing and altering the routing information and various active and passive attacks like Hello Flood Attack, Jelly Fish Attack, Sinkhole, Blackhole, Gray hole etc. in turn affecting the entire WSN network in terms of performance.
Limited Memory and Storage Space: A sensor is tiny device with limited amount of memory and storage space for code. In order to design efficient security mechanism, it is necessary to limit the code size.
1.4 Wireless Sensor Networks – Types Depending on the monitoring environment, the sensor nodes forming sensor networks can be deployed underwater, underground, land or even space. The following are the five major types of WSN network: a)
Terrestrial WSNs
b)
Underwater WSNs 7
c)
Underground WSNs
d)
Mobile WSNs
e)
Multimedia WSNs
Fig. 1.3. Types of Wireless Sensor Networks a)
Terrestrial WSNs: In terrestrial WSNs, thousands of sensor nodes are deployed either in ad hoc manner or pre-planned manner and are efficient to communicate with base stations. Structured mode takes into consideration optimal placement, grid placement or 2D/3D placement models whereas in unstructured mode, the nodes are installed in random fashion over target area. Energy is the primary issue with terrestrial WSNs and energy retention can be done via low duty cycle operations, delay minimization and energy efficient routing protocols.
Fig. 1.4. Terrestrial WSN 8
b)
Underwater WSNs : Underwater WSNs comprise of sensor nodes and sensororiented vehicles to uniformly operate underwater. Autonomous underwater vehicles are deployed for capturing sensed data from sensor nodes. The chief issues surrounding underwater WSNs are long propagation delay, quality of service, energy and sensor failures due to underwater obstacles.
Fig. 1.5. Underwater WSN c)
Underground WSNs : As compared to terrestrial WSNs, underground WSNs requires more cost in terms of deployment, maintenance and planning. Underground WSNs are deployed for monitoring underground conditions and communicate back to sink nodes above the ground. Sensor nodes are really difficult to charge and maintain, as nodes operate under the ground. Other issues surrounding underground WSNs are QoS wireless communications as signal loss and attenuation always occurs.
Fig. 1.6. Underground WSN 9
d)
Mobile WSNs : Mobile WSNs operate in physical environment for capturing real time data and are more versatile as compared to static sensor networks. Mobile WSNs are better in terms of area coverage, QoS, energy efficiency, better channel capacity etc.
Fig. 1.7. Mobile WSN e)
Multimedia WSNs : Multimedia WSNs perform the task of tracking and monitoring of events like imaging, audio, video etc. They are low cost sensor networks integrated with microphones and cameras and connect with other nodes in network via wireless communication for data compression, retrieval and correlation.
Fig. 1.8. Multimedia WSN
10
1.5
Wireless Sensor Networks- Classifications
WSNs are classified into the following different categories:
Static and Mobile Network: WSN network can be static or mobile. In static sensor network, all sensor nodes are static without any movement. Static sensors can be deployed in Environmental monitoring or surveillance. Mobile Sensor Network are more versatile and can be deployed in any scenario and cope with rapid topology changes. A Wireless Biosensor network for monitoring animals is Mobile Sensor Network.
Deterministic and Nondeterministic Network: In Deterministic Sensor Network, the positions of sensor nodes are pre-planned and are fixed once deployed. It is very challenging to deploy sensor nodes in pre-planned manner because of the harsh or hostile environments. Nondeterministic sensor networks are more scalable and flexible, but require higher control complexity.
Static-Sink and Mobile-Sink Network: In a static-sink network, the sink is static with a fixed position located close to or inside a sensing region. A static sink makes the network easy to control and administer, but it would cause a hotspot effect resulting in that sensor nodes close to data sink would die early and would interrupt normal network operation. In a mobile-sink network, the sink moves around the network to collect data from sensor nodes.
Single-Hop and Multi-Hop Network: In single-hop network, all sensor nodes transmit their sensed data directly to the sink node, which makes easy implementation of network control. This in turn, requires long range wireless communication which becomes costly in terms of energy and hardware and if the overall network size increases, energy will reduce and collisions will occur. In a Multi-Hop network, sensor nodes transmit their sensed data to the sink via intermediate nodes. Intermediate node perform the task of routing to forward the sensed data back to sink node.
11
Self-Reconfigurable and Non-Self-Configurable Network: In a non-selfconfigurable network, sensor nodes have no capability to organize themselves and rely on central controller to control each sensor node and collect information. In self-reconfigurable network, sensor nodes are able to organize autonomously and maintain the connectivity and path for transmission of data.
Homogenous and Heterogeneous Network: In a homogeneous network, all sensor nodes have same capabilities in terms of energy, computation and storage. In Heterogeneous network, sensor nodes are somewhat complicated as nodes are equipped with more processing and communication capabilities as normal nodes. Heterogeneous networks are better in terms of energy and overall life span.
1.6
Wireless Sensor Network Architecture: Protocol Stack
The protocol stack used by sensor nodes to communicate with base station is shown in figure 2.9. According to Akyildiz [2], the protocol stack of WSN is almost like traditional network protocol stack with multiple layers: Application, Transport, Network, Data link, Physical layer, Power management plane, mobility management plane and task management plane. [11] The primary responsibilities of physical layer are frequency selection, carrier frequency generation, signal detection, modulation and data encryption. Data link layer performs the tasks of data streams multiplexing, data frame detection, medium access and error control. Network layer is responsible for routing the information i.e. calculating the most efficient path of data transmission from source to destination. Network layer design in WSNs also consider energy efficiency, data-centric communication, data aggregation etc. Transport layer performs the task of data flow maintenance especially when data captured by sensor nodes is to be accessed via Internet or external communication networks.
12
The application layer does the task of presenting all the information to specific application and forwarding requests from application layer to lower layers in protocol stack. The software in application layer depends on sensor application. The power management plane is mainly responsible for minimizing power consumption in sensor nodes and waking up only those nodes required for packet transmission. The mobility management plane detects and registers movement of nodes to maintain the route of sender node to sink node. The task management plane balances and schedules the sensing tasks to the sensing field and only specific nodes are assigned the sensing tasks and rest other nodes can perform other tasks like routing and data aggregation. A network routing protocol for wireless sensor networks should be able to overcome all the necessary limitations and back points like Bandwidth, energy, electromagnetic wave propagation, congestion, collision and dynamic topology mobility.
Fig. 1.9. Protocol Stack of Wireless Sensor Networks
13
1.7
Routing in Wireless Sensor Networks
In general terms, Routing is regarded as process of discovering the path and transmission of data between source to sink node. In Wireless Sensor Networks, data routing is performed at network layer. Routing in WSNs is regarded as highly challenging task due to following inherent characteristics which distinguishes WSNs with other types of wireless networks like Mobile Adhoc Networks (MANETS), Cellular or Home based Wi-Fi or WiMAX communications: [12, 13, 14]
First, due to large number of sensor nodes deployment in real-world, which becomes somewhat hard to design a global addressing scheme for every node involved in sensor network as the overhead of unique ID maintenance is high.
Sensor nodes are deployed in ad hoc manner so nodes have to work via selforganization, but with changing environmental conditions and limited energy of nodes with dynamic changing topology, this becomes really challenging.
Sensor nodes have limited energy, storage and processing capabilities. So, careful resource management is required.
Sensor applications varies from conditions to conditions. In some conditions, sensor nodes remain stationary but in some applications, sensor nodes may be allowed to move and change location.
Position of sensor nodes is very important as data collection is normally based on location. But currently, it is not practically possible to integrate GPS (Global Positioning System) in sensor nodes. Methods based on triangulation facilitates sensor nodes to approximate their positions via radio signals to few points.
Lastly, the data collected by sensors deployed in WSNs are of similar nature, which leads to high probability of data redundancy. Redundancy needs to be tacked by routing protocols to improvise energy and bandwidth.
Considering these above characteristics, the routing protocol must be simple, robust, energy efficient to maintain overall efficiency in network.
14
1.7.1
Challenges connected to Routing in Wireless Sensor Networks
Considering different real-time application scenarios, architectures and goals, Wireless sensor networks face different routing challenges and the performance of routing protocol entirely depends on architectural model. The following points highlight the various challenges connected to routing in WSN [1, 5, 6, 12]:
Deployment of Sensor Nodes: Sensor nodes deployment entirely depends on applications and have straight effect on protocol performance. The deployment can be either deterministic or self-organizing. In deterministic deployments, the nodes are installed as per plans and routing is performed via pre-determined paths. But in case of self-organizing, the nodes are installed in random manner. The position of sink node or cluster head is highly important to maintain energy efficiency and performance in network.
Network Dynamic: Depending on applications, sensor nodes remain stationary but sometimes sensor nodes change the position and in that case topology become dynamic. Route stability is highly important to conserve energy and maintain overall QoS in network.
Energy considerations: The most important factor to consider for setting up the route between source to destination is Energy Efficiency. As the transmission power of wireless radio is directly proportional to distance squared or high order in lieu of obstacles, multi-hop routing will be beneficial in terms of energy efficiency because of effective topology management and medium access control.
Data delivery models: Data delivery models to sink node can be continuous, event-driven, query-driven or hybrid and all depends on sensor application operation. In continuous delivery model, every sensor in network sends data at periodic intervals. In event-driven and query-driven models, data transmission only happens on specific event triggering. In case of hybrid data delivery model, it makes use of combination of continuous, event-driven and query-driven data 15
delivery models. The efficiency of routing protocol highly depends on selecting the effective data delivery model to utilize less energy and provide QoS in overall network.
Data Aggregation: Sensor nodes generate same type of data which can lead to redundancy. It is regarded as effective collaboration of sensed from varied sources via using functions like suppression (Duplicates elimination), min, max and average. Routing protocols should carefully perform data aggregation to maintain overall transmission speed.
1.7.2 Classifications of Routing Algorithms/Protocols Considering various issues and challenges of routing in sensor communications, various novel algorithms/protocols are proposed to for efficient routing in sensor networks. Figure 2.10 demonstrates the types of routing algorithms. Routing algorithms can be classified into following two major categories: 1)
Adaptive Algorithm
2)
Nonadapative Algorithm
Nonadaptive algorithms (or Static Algorithms) don‟t adapt themselves to any sorts of changes in overall network. Adaptive algorithms (or Dynamic Algorithms) change according to the topology and traffic load in overall network.
Fig. 1.10. Classifications of Routing Algorithms 16
Nonadaptive Routing Algorithms Nonadaptive routing algorithms don‟t take routing decision on basis of measurements and estimates of current traffic and topology. The routing path from node to node is determined in advanced, off-line and downloaded to routers when booting of network is performed. Non-adaptive routing is also known as static routing. In terms of static routing, any change in topology or loss of link, the node is not compensated, which means either packet transmission would be halted or repaired or restarted. It can be classified into following two types:
Shortest Path Routing
Flooding
Shortest Path Routing Shortest path routing algorithm determines the shortest path between nodes on network and select the shortest route. In general sense, the labels on the link can be computed as a function of distance, communication cost, delay, average traffic, bandwidth, queue length etc. The node to transmit has to determine the shortest distance to next node overall maintaining shortest path from source node to sink node. Flooding In flooding, every node in network transmits packets to all neighbouring nodes creating huge amounts of duplicate or redundant packets. In case of WSNs, flooding can be used when all the messages transmitted by a station are received by all other stations within their communication range. Flooding creates overall congestion, network load and also utilizes much energy of each and every node in the network. Adaptive Routing Algorithms Most of wireless communication networks makes use of Adaptive routing algorithms. The Adaptive routing algorithms (Dynamic Routing Algorithms) change their 17
respective routing decision with regard to any sort of change in network topology and traffic load. Adaptive routing mainly defines the ability of the system, in calculating the optimized routes from source to destination and to modify the path the route takes through the system with regard to changing conditions. Adaptive routing is basically designed to allow as many as routes as possible to cope up with packet loss, in response to any sort of change in the network. Adaptive routing is utilized in networking to describe the ability of network to make changes in route discovery, path establishment and data transfer between nodes. Adaptive routing is broadly classified into following three main categories: Data Centric, Hierarchical and Location based routing protocols. [5, 6, 14]
Data Centric Routing (Flat Routing): Depending on the applications, where tons of sensor nodes are deployed, it becomes really challenging to assign Global Identifiers to each node [15, 16]. With the utter lack of global identifications in addition to random deployment of nodes, it becomes really difficult to query a specific sensor node. This leads to redundancy in entire network. This problem has led to development of Data-Centric routing. In datacentric routing, the sink sends queries to certain regions and wait for data to be received back from sensors located in selected regions. Data Centric protocols are: SPIN, Directed Diffusion, Rumor Routing, GBR, MCFA, CADR, COUGAR, ACQUIRE, EAR.
Hierarchical Routing: Hierarchical Routing or Cluster based routing [14] are regarded as specialized techniques providing specific advantages in terms of scalability and efficient communication to wireless sensor networks. As sensor nodes are not capable of long-haul communication, single gateway architecture is not scalable for large set of sensors. Network clustering has been proposed in various routing protocols to balance additional traffic load and cover large area 18
without depreciating the service quality. In hierarchical based architecture, the nodes with higher energy are used to process and transmit the information while low energy nodes perform the task of sensing in the proximity of the target which leads to cluster creation and cluster head performs specific tasks making the overall WSN network scalable and energy efficient. The main objective of hierarchical routing is to maintain energy efficiency of nodes by using multi-hop communication in cluster and perform data aggregation to reduce transmitted packets. It is basically two-layer routing in which one layer makes the selection of cluster heads and other layer performs routing process. Hierarchical protocols are: LEACH, TEEN & APTEEN, MECN& SMECN, SOP, HPAR, VGA, SENSOR Aggregate and TIDD.
Location based protocols: Location based protocols [14] are required to calculate the distance between two sensor nodes to determine the energy consumption. Location based protocols use the position information to spread the data to desired regions rather than entire network. As, there is no utilization of addressing scheme as in IP-addressed networks, location information can be used to route data from source node to destination node in energy efficient way. Location based protocols are GAF, GEAR, SPAN, MFR & GEDIR and GOAFR.
Swarm Intelligence and Ant Colony Optimization 1.8
Swarm and Swarm Intelligence (SI)
In the past few decades, the coordinated and complex behaviour of swarms have also been a source of interest for biologists and also computer scientists because of amazing efficiency in solving complex problems. In 1989, the concept of “Swarm Intelligence” was proposed by G. Beni and J. Wang [32]. Particle Swarm Optimization (PSO) based on Birds and Fish Swam (FS) based on fishes are regarded as the most important examples of coordinated behaviour that emerges without any 19
need of centralized control [29]. Social insect colonies like Ants, Termites demonstrate complex problem solving skills arising from actions and interactions of nonsophisticated individuals. Natural systems solve multifaceted problems via simple rules, and exhibit organized, complex and intelligent behaviour [30, 31]. Natural process control systems are highly adaptive, evolutionary, distributed (decentralized), reactive and aware of environment. Considering the foundations of millions of years of evolution, biological based systems and processes have following main characteristics:
Highly adaptive to dynamically changing environmental conditions.
Robust and compatible to failures mainly caused by internal or external factors.
Ability to attain complex behaviors on the basis of limited set of basic rules.
Ability to learn and evolve when new conditions are applied.
Self-organization ability in fully-distributed fashion, to achieve efficient equilibrium.
Best management of constrained resources.
Surviving capability in harsh environments.
Swarm Intelligence, is the collaborative intelligence of groups of simple individuals, called agents. In simple terms, Swarm Intelligence [26, 27, 28, 41] is regarded as collective behaviour emerged from social insects operating under few defined rules and regulations. With Swarm Intelligence, the developed algorithms have to be highly flexible to adapt internal and external changes and should be highly robust and scalable where individuals fail to be decentralized and self-organized. Swarm Intelligence lays foundation on following two principles: I)
Self-Organization
II)
Stigmergy
20
Self-Organization Bonabeau et. al, in Swarm Intelligence, 1999 defined the concept of Self-Organization as, Self-Organization is a set of dynamical mechanisms whereby structures appear at the global level of a system from interactions of its lower-level components” Self-organization based system is characterized by following three main parameters:
Structure: emerges from a homogeneous start-up state. Example: Ants foraging trail, Nest Architecture.
Multi-stability: coexistence of many stable states. Example: Ants exploits only one of two identical food sources.
State Transitions: dramatic change of system behavior. Example: Termites operate by moving from non-coordinated to coordinated stage only if their destiny is higher than a threshold value.
Self-Organization has foundation from following four major components: a)
Positive Feedback: It is a simple behavioural “rules of thumb” which lead to structures creation. Examples of positive feedback are: Recruitment and Reinforcement. In terms of ACO, recruitment and reinforcement is “Trail Laying” whereas in PSO is “Birds Flocking”.
b)
Negative Feedback: It counterbalances positive feedback and provides backbone for stabilizing collective pattern. Talking of ACO foraging, negative feedback is required at the stage of food source finish, crowding or completion of food source.
c)
Fluctuations: It includes random walks, errors, random task-switching. Randomness is most important aspect for emergent structures in terms of new food discovery, locating unexplored food locations and recruitment of nest mates to discovered sources of food. 21
d)
Multiple Interactions: It occurs in swarms as all the agents scatter information with each other coming from other swarms so that data spread across entire network and swarms can have self-organization movement from nests to food sources.
Self-Organization is very well achieved by swarms via two sorts of communication: Direct Communication and Indirect Communication. Stigmergy The term “Stigmergy” [38, 39] was introduced by Grassé. The word “Stigmergy” comprises of two words: Stigma meaning Sting and Ergon meaning work and in turn means Stimulation by work. In addition to Self-organization, stigmergy is also an important term which defines work coordination and also work doesn‟t depend solely on workers. In insects, stigmergy is closely observed. In terms of ants, ants exchange the path from nest to food source by splitting pheromone on the way so that other ants follow the trail and lay the foundation of food transport from food source to nest. Taking stigmergy with respect to computer science, stigmergy can be applied via Ant Colony Optimization, in which ACO technique searches for solutions to various complex problems via “Virtual Pheromones” and choosing the best path as solution. Stigmergy is of following types: 1)
Sign-based Stigmergy: Pheromone based trail following by ants from nest to food source.
2)
Sematactonics: nest building by termites.
3)
Quantitative: Trail-laying and following behavior of ants.
4)
Quantitative: Nest building by social wasps.
Swarm Intelligence based algorithms and techniques are used in varied optimizations tasks and research areas. Swarm Intelligence principles are successfully applied in 22
various problem solving domains including function optimization problems, determining efficient routes, image and data processing [33, 34]. To understand, model and simulate the broad behaviors arising from Swarm Agents, the following are some of the principles of Swam Intelligence defined by Millonas in 1999: [24, 25, 40]
Principle of Proximity: The swarm are regarded as highly capable agents of performing simple computations relating to the surrounding environment. The computation is regarded as direct response to behavior with regard to environmental variations. Depending on the agent complexity, responses vary from situation to situation like searching for food in live changing environment considering all sorts of natural and path obstacles and nest construction by ants.
Principle of Quality: In addition to solving complex computation problems, a swarm should response to various quality factors like food, safety in harsh environment and adapting to changes in environment.
Principle of diverse response: With regard to resources, concentration should be in narrow areas. The search of food by agents should be such organized that every agent should be protected against all sorts of changing conditions in environment.
Principle of Stability and Adaptability: All sorts of environmental problems should be efficiently tackled by swarms and should save all resources in terms of costs and energy in search of food resource.
The most popular Swarm Intelligence based algorithms [25, 29, 30, 35, 36] are Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Artificial Bee Colony (ABC), Bacterial Foraging, Fish Swarm, Cuckoo Search, Bat Algorithm, Pigeon Swarm, Elephant Swarm, Cockroach Swarm, Firefly Algorithm, Cat Swarm, Glow-worm Swarm Algorithm, Flower Pollination Algorithm and many more. Human beings when possibly stuck in hard complex computation problems, always move towards Swarm Intelligence based algorithms. On Swarm Intelligence grounds, 23
social insects (agents) play a dramatic role in problem solving because of their natural intelligence. Social Insects solve the problem individually or via self-cooperation. Ants based swarm search for food highly demonstrates the ability to determine the shortest path from source to nest. A flock of birds fly across of sky without a single collision lay a strong foundation for telecommunications and networks based problems of collision avoidance. A school of fish swim together. This kind of collective behavior known as swam intelligence can be applied to tons of engineering issues and problems to determine the solutions to model the behavior via mathematical analysis and simulation. With the use of ants and other social swarms as models, the development of software agents is possible to find solution to complex problem like rerouting the traffic in busy telecommunication network [37]. In this thesis, we focus primarily on the most popular Swarm Intelligence technique utilized by various researchers to solve various issues of Wireless Sensor Networks, namely, Ant Colony Optimization.
1.9
Ant Colony Optimization (ACO)
1.9.1 Introduction The concept of “ANT COLONY OPTIMIZATION” [42, 43, 44, 45, 46, 47, 48] was coined by Marco Dorigo and colleagues for finding solutions to hard combinatorial optimization (CO) problems. It is probabilistic technique and ACO algorithm is member of ant colony algorithms family in the broad area of Swarm Intelligence which consists of some meta-heuristic optimizations. The concept of Ant Colony Optimization was basically developed and conceptualized by Marco Dorigo [44] in his Ph.D Thesis as first algorithm whose basic aim was to find the shortest path in the graph by conceptualizing the problem based on ants finding food from their respective colonies to the food source. The proceeding section outlines ants in nature, Ants Stigmergic behavior, Real Ants v/s Artificial Ants, ACO Metaheuristic and ACO based Algorithms- Ant System (AS), Ant Colony System (ACS) and MAX-MIN Ant System (MMAS).
24
1.9.2
Ants in Nature
Ants are able to survive tons of dynamic changing environments, climates. The secret behind the success of survivability of ants against all sorts of harsh environments is Sociality [50]. Ants colonies [48, 49], more specifically social insects are distributed systems that, in spite of simplicity of individuals, present a highly structured social organization. As a result, they live in organized societal setup made of agents that cooperate, communicate and divide the tasks equally for solving complex problems. The main backbone behind Ants coordination is Self-Organization principle which demonstrates impressive abilities in doing varied tasks like shortest path searching, nest building and locating food source and laying the trail of pheromone for food transport from source to nest. Biologists have demonstrated that Stigmergic, indirectcommunication which demonstrates how a social agent achieve self-organization. Ants being highly efficient, hardworking and highly adaptable to changing conditions leads to applicability of Ants based algorithms in different engineering disciplines. 1.9.3
Ants Stigmergic behavior
Stigmergy, being a biological term is related with insect depicting swarm behavior and describes a model supporting environmental communication separately from artefacts or agents. Ants, being social insects like many other swarm agents, communicate via chemical substance known as pheromones which lays the foundation of direction and intensity of food. The word “pheromone” was coined by P.Karlson and M.Lusher in 1959, where pherein means transport and hormone means stimulate [51]. Insects use different types of pheromones. Talking of ants, two main types of pheromone which are used by them are: Alert Pheromone- which alerts the ants to alert regarding nearby dangerous predators to protect their respective ant colonies; Food Trail Pheromone- where other ants follow the trail of pheromone from start (Nest) to Destination (Food Source). Ants which make the shortest path from source to food and back to nest will attract other ants to follow as seen in figure 3.1. Positive feedback process is example of self-organization behavior of ants to determine the probability of selecting the shortest path and the level of pheromone increases as more ants traverse the same path. 25
Fig. 1.11. Ants stigmergic behavior After transferring the whole food from source to nest, no pheromone trail is laid by ants returning back and pheromone being chemical substance gets evaporated with passage of time. This sort of negative feedback enables ants to handle the environmental changes. When the path is obstructed by any sort of obstacles, the ants search for new paths. Such sort of trail-laying and following behavior is known as Stigmergy and usually regarded as indirect communication in which ants automatically change with environment and responds to new environment automatically. [30, 34] 1.9.4
Real Ants v/s. Artificial Ants
Real Ants are those which use natural phenomenon to discover optimal paths, search food in the environment via probability, build nests and lay pheromone trail. Artificial Ants is regarded as developing a nature inspired algorithm based on real ants. Natural phenomenon is highly constrained and can be only done via observations and experiments. Development of nature inspired algorithm is totally dependent on one‟s understanding towards working of algorithm, mathematical computations deriving the behavior and technology to simulate the model. Ant Colony Optimization (ACO), is completely based on real ants social behavior, but it is not possible to design artificial ants as the same replica. Modeling is regarded as interface between natural behavior understanding and development of artificial systems. Technically, researchers start with natural phenomenon observation, develop a nature-inspired model of then and then artificial system of it without any sort of constraints. 26
Table 1.2. Real Ants V/s Artificial Ants. Basis of Difference Depositing Behavior Pheromone
Real Ants
Artificial Ants
of Real ants lay a trail of pheromone while moving from source to nest and vice versa.
Pheromone is laid on return way after a candidate solution is developed and thoroughly tested.
Updating of Pheromone
Trail of pheromone on the path is updated depending on food quality and quantity.
After construction of food source path, the pheromone trail is updated on return way with the amount that is inversely proportional to path length stored in ants memory
Memory
No Memory at all
While traversing the path from nest to food source, the path is stored in memory to be used for retreat. Memory is also used to determine the quantity of pheromone deposited on their return back to nest.
Return Path Model
Ants use the pheromone deposited on their forward way to retrace the return path back to their nest.
As artificial ants don‟t make use of pheromone trail deposited on search of path, they make use of memory which has stored path to traverse back to nest.
Pheromone Evaporation
Pheromone being It evaporates exponentially chemical substance making it more significant evaporates after period of for the convergence. time once the food is finished or other shortest route is determined.
Table 1.2 defines the possible differences between real ants and artificial ants. “Memory” is one of the major difference between real and artificial ants. This memory lays the foundation for artificial ant to implement a large number of effective behaviors to build solutions to complex problems.
27
1.9.5 Ant Colony Optimization Metaheuristic Optimization Technique was proposed by Deneubourg. The reason behind the development of this technique was to explain the foraging behavior of ants. In Ant Colony Optimization (ACO), ants basically build the optimized path among various paths to transfer food to the nest and also alert other ants with information of the food source. Ant Colony Optimization is regarded as Metaheuristic Technique for solving combinatorial optimization problems and was given by Marco Dorigo [53] [47] [55]. The development behind ACO technique was the inspiration from real ants. Ants live in colonies and their main aim is colony survival rather than individual interest. The main idea behind the development of ACO technique is Ants Forging Behavior i.e. in what scenario Ants are able to determine most efficient and shortest distance between nest and food source. Metaheuristic is defined as set of algorithms [49] that define a set of heuristic methods applied to widest range of issues in real world. It is regarded as general purpose algorithm based framework which can be utilized towards variety of optimization problems with limited modifications. Ant Colony Optimization is regarded as pure meta-heuristic technique which can solve wide problems and lead to various optimizations in fields of science and engineering. Researchers these days are using ACO techniques towards wide range of applications of computer science and solving even various types of NP-HARD problems. The basic inspiration behind the development of ACO was foraging behavior of real ants [52]. When searching for food, ants basically explore the area surrounding their nests in random manner. As soon as an ant finds a food source, it determines both the quantity and quality of food and picks up some of the food back to the nest. During the return of the ant from the food source to the nest, the ant deposits a chemical called Pheromone on the ground which depends on the quantity and quality of food, which acts as navigation path for the rest of ants from the nest to the food source. 28
1.9.6
Mathematical Model of Ant Colony Optimization
The Ant Colony Optimization [56] [57] mathematical model was applied to find solution to “Travelling Salesman Problem” in which the main aim is to discover shortest path from one source to destination point among lots of paths which are available. The algorithm is simple and based on a set of ants, where each ant is making a move from one place to another in search of food. The main characteristic of this algorithm is that pheromone values are updated by all the ants which have completed the tour. The update with regard to Pheromone for Tij for joining places i and j is as follows: Each move is based on local probability value as:
[ ij (t)] *[ij ] , j Nlk k Pij (t ) ik [ il (t)] *[il ] 0 , j Nlk
… (1)
After the ants in algorithm ends the tours, the trail levels are updated based on the tour length of previous solution and evaporation on basis of following formulas:
ij (t ) ij (t 1) ij
… (2)
m
ij ijk
… (3)
Q if ant k uses arc (i j ) in its tour Lk 0 otherwise
… (4)
k 1
k ij
This iteration process on until a certain terminal condition: a certain number of iterations have been attained, a fixed amount of CPU time has elapsed, or solution quality has been achieved.
29
1.9.7 Components of Ant Colony Optimization (ACO) The following points highlight the basic models of Ant Colony Optimization:
A set of concurrent computation agents (ANTS)
Each ant moves based on stochastic local decision based on: o Trails (globally affected) o Attractiveness (locally affected)
Global mechanisms o Trail evaporation o Daemon actions
Ants iteratively constructs the solution by local choices from state i to state j
At each step σ, ant k computes a set of feasible expansions Ak (i) from its state.
Probability of moving from state i to state j pik j depends on: o Attractiveness nij of the move o Trail level tij of the move
Fig. 1.12. Ant Colony Optimization
30
1.9.8
Ant Colony Optimization Algorithms
The four most successful Ant Colony Optimization Algorithms are as follows: [54, 56, 117] 1)
Ant System
2)
Ant Colony System
3)
MAX-MIN Ant System (MMAS)
4)
Ant Lion Algorithm
1.9.8.1 Ant System Algorithm The first algorithm which was developed under Ant Colony Optimization [54] [55] [56] [57] was Ant System and the motive behind its development was to find solution to “Travelling Salesman Problem” in which the main aim is to discover shortest path from source to destination point among lots of paths which are available. The algorithm is simple and based on a set of ants, where each ant is making a move from one place to another in search of food. The main objective of Ant System Algorithm is all the ants update the pheromone values after tour is completed [57]. The update with regard to Pheromone for Tij for joining places i and j is as follows:
Each move is based on local probability value as:
i j i j if (i j )tabuk pik j i l i j (i l )tabuk 0 otherwise
… (5)
Trail Levels are updated based on the tour length of previous solution and evaporation:
i j (t ) i j (t 1) i
… (6)
j
m
i j ik j
… (7)
k 1
k i j
Q if ant k uses arc (i j ) in its tour Lk 0 otherwise 31
… (8)
Algorithm for Ant System 1)
Initialization: Initialize Tij and nij values
2)
Construction foreach ant k (in state i) do: repeat choose the state j to move to (with prob.) Append the chosen move to tabuk Until ant k has completed its solution
3)
Trial Update For each ant move (i j) do: compute ΔT (i j) update trial matrix
4)
Termination if not end of test, go to step 2
The following are regarded as extensions to Ant System:
Elitist Ant System (the elite updates trails at each cycle)
Ant Colony System
Max Min Ant System (Max and Min Pheromone)
1.9.8.2 Ant Colony System (ACS) The Algorithm “ANT COLONY SYSTEM” is regarded as improvement to Ant System Algorithm in several ways. The Ant Colony System is regarded as the first algorithm which is based on the behavior of real ants. The main aim behind the development of this algorithm is to find solutions to more complex optimization problems [58]. The algorithm was developed by Dorigo and Gambardella in 1997. The development of ACS was done to show the potential of using artificial pheromone and artificial ants to find better solutions to complex optimization problems [57]. The enhancement which has been made by Ant Colony System is via the introduction of local pheromone update after every construction step by all ants. 32
The three basic changes which are there in ACS as compared from Ant System are as follows:
Pheromone, trail update (Best Updates)
In ACS once all the ants have computed their tour (i.e. at the end of each iteration) AS updates the pheromone trail using all the solution provided by the ant colony. Each edge belonging to one of the completed solutions is modified by an amount of pheromone of the entire system evaporates and the process of construction and update is iterated. On the contrary, in ACS only the best solution computed since the beginning of the computation is used to globally update the pheromone. As was the case in Ant System, global updating is intended to increase the attractiveness of promising route but ACS mechanism is more effective since it avoids long convergence time by directly concentrate the search in a neighborhood of the best tour found up to the current iteration of the algorithm. [26]
State transition rule (choice of exploration v/s exploitation)
During the construction of a new solution the state transition rule is the phase where the ant decided which is the next state to move on. In ACS, a new state transition rule is introduced called pseudo-random-proportional. It is basically a compromise between the pseudo-random state choice rule typically used in Q-Learning [59] and the randomproportional action choice rule typically used in Ant System. A choice is made among the best transition and probabilistically among others.
max (i s )tabuk i j i j if q q0 (exploit) s proportional to weighted values otherwise (explore) q: random value such that 0 q 1
… (9)
Hybridization
As, ACS was being developed to find solutions to complex travelling salesman problems. For this purpose, ACS incorporates advanced data structure known as “CANDIDATE LIST” [59]. An ant in ACS first makes use of candidate list along with 33
state transition rules to choose the area to move. If none of the area in candidate list can be visited by the ant, the ant chooses the nearest area using heuristic value and this choosing has been improved by optimization via hybridization. ANTS algorithm is regarded as enhancement to Ant Colony System (ACS) which has two mechanisms:
Attractiveness
Trail Update
1.9.8.3 MAX-MIN Ant System MAX-MIN [60] was developed by Stutzle and Hoos in 2000 with significant improvements to original Ant System in following ways: The best ant adds the pheromone trails and the value of the pheromone is bound. It can be represented with following mathematical formula:
(1 ) ij ij
ij
… (10)
ij
if (i j) forms the efficient tour. Where Δτijbest=1/Lbest if the best ant used edge (i,j) in its tour, Δτijbest =0 otherwise, where Lbest is the length of the tour of the best ant. As in ACS, Lbest may be set (subject to the algorithm designer decision) either to Lib or to Lbs , or to a combination of both. 1.9.8.4 Ant Lion Optimizer Ant Lion Optimizer [61] algorithm was proposed by Seyedali Mairjalili in 2015 for solving constrained problems with diverse search spaces. The algorithm is being developed on natural behavior of hunting mechanisms of ant lions which comprise of five main steps: Ants Random Walk, Traps Development; Entrapment of ants in traps; Preys Catching; Re-building of Traps. 34
Ant Lion Optimizer Algorithm Initialization of the first population of ants and antlions randomly Determining of fitness of ants and antlions Calculation of best antlions and assuming Best AntLion as elite (Determined optimum) While the end criterion is not satisfied For every ant Select an antlion using Roulette wheel Update c and d using Equations Create a random walk and normalize it using equations Update the position of ant End for Calculate the fitness of all ants Replace an antlion with its corresponding ant it if becomes fitter End while Return elite. 1.9.9 Ant Colony Optimization- Working and Algorithm Ant Colony Optimization technique [44, 45, 46, 47, 48, 26] is based on ants i.e. how ant colonies find the efficient path between nest and food source. In search of food, ants roam randomly in the environment. On location of the food source, ant‟s first return back to their nest by laying a trail of chemical substance called “Pheromone” in their path. Pheromone lays the foundation for communication medium for other ants to follow the way and go to the food source. When other ants follow the path, the quantity of pheromone increases on that particular path. The rich the quantity of pheromone along the path, the more likely is that other ants will detect and follow the path. In other words, ants follow that path which is marked by strongest pheromone quantity. As pheromone evaporates over time, which in turn reduces its attractive strength? The longer the time taken by ant to travel the path from food source to nest, the quicker the pheromone will evaporate. So, the path should be shorter so that the 35
active strength of pheromone is maintained and ants can easily transfer the food from source to nest. So, in turn of this policy the shortest path will naturally emerge. Ant Colony Optimization Algorithm (ACO) Algorithm Initialize Parameters Initialize pheromone trails Create ants While Stopping criteria is not reached do Let all ants construct their solution Update pheromone trails Allow Daemon Actions End while Ant Colony Optimization Algorithm
1.10 Suitability of Ant Colony Optimization (ACO) based approach for Developing Energy Efficient Routing Protocols for Wireless Sensor Networks Wireless Sensor Networks have recently gained popularity amount researchers due to recent advancements in Wireless Communications, Digital Electronics and MEMS 36
technology and is suitable for those situations where pre-deployment of nodes planning is not possible at all. Due to rapidly changing topology and environmental conditions, there are lots of challenges which can lead to implementation issues. Some of the known challenges are: (1) Effective distribution and forwarding of data among nodes. (2) Efficient load balancing of traffic load among all the nodes. (3) Achieving energy efficiency and overall improvement in network lifetime of sensor nodes. (4) Achieving energy efficient routing among nodes- with only those nodes active which are required, rest in sleep mode. (5) Congestion avoidance and effective throughput by maintaining packet delivery ratio and reducing routing overhead. (6) Maintaining fault tolerance and scalability in entire network. The following points highlight the Suitability of ACO based Routing Protocols for Wireless Sensor Networks: [66] [20] 1)
As ACO algorithms are fully distributed, so failure rates are reduced to large extent in sensor nodes communications.
2)
Simple operations can be performed in each and every node for routing of packets among nodes and back to sink node.
3)
Autonomous integration of ants, and the algorithms are based on Agent‟s Synchronous.
4)
ACO algorithms have capability of Self-Organizing which is very important as sensor nodes, when deployed randomly have to fully robust, scalable and fault tolerant. ACO algorithms makes WSN networks fully self-organization compliant.
5)
ACO algorithms are very well suited to adapt to all kinds of changes in realworld topology and increase in number of nodes and packet traffic.
6)
ACO algorithms solve complex CO problems, making them well suitable for highly complex situations when sensor nodes are deployed especially in RealTime Monitoring, Production and Military based Battlefield‟s monitoring.
Various energy efficient routing protocols are designed by various researchers to reduce routing overhead, maintain security and energy efficiency in the entire 37
network. But, every protocol has its own pros and cons and some are not effective when coming to real-time dynamic environment operations. Because of the efficient design and operation features of ACO-based algorithms and edge over solving complex computational problems, [62, 63, 64, 65], ACO based routing protocols provide the best optimized solutions to combat almost every challenge in effective operation of wireless sensor networks. The focus of our research is to overcome the limitation of Energy Consumption by proposing a Novel Energy Efficient routing protocol based on Ant Colony Optimization to significantly reduce routing overhead, improvise throughput, packet delivery ratio and End-to-End delay in Wireless Sensor Networks.
1.11 ACO Based Routing Protocols for Wireless Sensor Networks In this part, various ACO based routing protocols are comprehensively discussed [65, 66, 67, 68, 69, 70]: 1.11.1 Sensor Driven Cost-Aware Ant Routing (SC) The main problem surrounding all Basic Ant Routing Algorithms is that all the forwarding ants normally consume lots of time to locate the destination, even when a tabu list is being utilized (i.e. Repeating nodes are not included). This situation usually occurs when ants primarily don‟t have any idea regarding the exact destination. Only when the destination is located, the links are traversed along with certain probabilities of link exchange. In SC Routing [71], the routing performance is improvised, it is assumed that forward ants are equipped with sensors to locate the best destination for food at the initial process of routing. In addition to smart sensing ability of ants, each node stores the probability distribution and every node estimates and stores the cost to the destination from neighbouring nodes. It suffers from redundant data when obstacle arise in path leading to sensing errors.
38
SC Algorithm:
1.11.2 Energy Efficient Ant Based Routing (EEABR) Energy Efficient Ant Based Routing (EEABR) algorithm, proposed by T. Camilo et al [22] is an improvised routing protocol based on Ant Colony Optimization (ACO) metaheuristic. The protocol was designed with an objective to enhance sensor nodes energy by reducing communication overhead in discovering the paths from source to destination. The protocol adds new functionalities in pheromone tables updation of sensor nodes. Algorithm 1)
In EEABR routing protocol, at regular interval period of time, from each network node, a forward ant is launched to determine a path from nest to food source. The identifier of every visited node is saved in memory and carried 39
forward by ant. Each network node has routing table with N entries, one for each possible solution, and destination is one of the entry in nodes routing table. 2)
At every node, the ant selects the next hop using the same ACO metaheuristic probabilistic rule.
3)
When the forward ant reaches the food destination, it is transmitted back to proceeding ant, whose main task is to update the pheromone trail of the path used by forward ant to reach from nest to source and also stored in memory.
4)
The destination node computes the amount of pheromone trail that the ant will drop during the journey, before backward ant starts the journey.
5)
When the node, receives the backward ant coming from neighbouring node, it updates the routing table.
6)
When the backward ant reaches the nest, the actual path is determined by other ants to follow.
Simulation of EEABR with BABR (Basic Ant Based routing algorithm) and IABR (Improvised Ant-Based Routing Algorithm) is done on NS-2 simulator on varied parameters like Average Energy, Minimum Energy, Standard Deviation and Energy Efficiency and overall EEABR performs much better as compared to other two routing protocols. The only drawback of EEABR is lack of QoS and somewhat delay in packet delivery. 1.11.3 Flooded Forward Ant Routing (FF) Flooded Forward Ant Routing (FF) [71] was developed to overcome the shortcomings of misguiding paths due to obstacles in SC protocol even when ants are equipped with sensors. When the exact destination is unknown at the beginning by ant and even the cost cannot be determined, SC protocol was reduced to Basic Ant Routing and still the problem of unknown wandering around the network by ant to find the destination exist. In that case, FF protocol was introduced to remove the problem. FF protocol exploits the network via broadcast channel of WSN which means FF protocol makes use of Broadcast method of sensor networks to route the network
40
packets from source to destination. The objective is to flood forward ants to the destination. If the food search is successful, forward ants will direct backward ants to traverse backwards to the source. Multiple paths are updated by one flooding phase and probabilities are updated in the same manner as in Basic Ant Routing Protocol.
Algorithm for Flooded Forward Ant Routing 1.11.4 Flooded Piggyback Ant Routing (FP) In flooded Piggyback Ant Routing (FP) [71], a novel specimen of ants i.e. Data Ants was introduced. The forward list is carried by FP. In FP protocol, forward ants and data ants are combined via constrained flooding to route data packets and search for energy efficient paths in the network. FP protocol was compared with SC, FF and Basic ACO routing protocols in RMASE (Routing Modeling Application Simulation Environment) simulator. Results showed FP is not an energy efficient routing protocol. FF protocol is efficient in reducing delay and SC remains highly energy efficient routing protocol among FF, FP and Basic ACO routing protocol. 41
Algorithm for Flooded Piggyback Ant Routing 1.11.5 Energy-Delay Ant Based (E-D Ants) Energy-Delay Ant Based (E-D Ants) [73] was proposed by Wen et. al (2008). E&D ants is a reactive routing protocol being based on ant algorithms for performing varied routing operations. E-D Ants Protocol is based on Energy*Delay metrics to enhance network lifetime and minimize propagation delay by making use of a novel variation of Reinforcement Learning (RL). The Mathematical expression of E-D Ants Protocol is: g(t) = min (Energy * Delay) The protocol works on Iterative generation and unicast transmission of multiple forward ants to minimize energy and delay like AntNet Protocol. In this protocol, every ant stores the residual energy level and hop delay experience in its stack moving from node to node. E-D Ants Routing protocol was simulated in OPNET Simulator using 50 sensor nodes in area of 100x100 m and compared with two routing protocols: AntNet and AntChain on basis of Energy Efficiency, Delay and Routing Overhead. The results showed E-D Ants Protocol is almost 150% efficient as compared to other two 42
protocols. E-D Ants protocol is also efficient routing protocol in determining optimal paths from source to destination. 1.11.6 Ant Colony Based Reinforcement Learning Algorithm (AR and IAR) Adaptive Routing (AR) and Improved Adaptive Routing (IAR), proposed by Ghasemaghaei et. al (2007) uses probability distribution like other Ant-Colony based routing protocols in finding optimal paths from source to destination. The only difference between AR and IAR with other ACO based routing protocols is the use of reinforcement learning algorithm by backward ants to get efficient routing path from source to destination. In AR and IAR [74, 75], two types of ants are deployed: 1)
Forward ant (Fant)- travelling from source node (s) to destination node (d)
2)
Backward ant (Bant): which is generated by Fant when Fant reaches the destination d.
The backward ant gets back to sink node via information supplied by forward ant. But backward ant makes use of reinforcement learning method to get better and most optimal route as compared to the route being chosen by forward ant and updates the routing table of sensor nodes visited during reverse journey. AR and IAR algorithms were simulated on Java based simulator using 7x7 sensor node grid for 200 seconds. AR and IAR algorithms are compared with 4 Routing Algorithms: Basic Ant Routing, SC Ant Routing, FF and FP Routing Algorithm on parameters like Latency, Energy Consumption, Success Rates. Simulation results showed AR and IAR much efficient in every parameter as compared to other 4 routing protocols.
43
Algorithm for Forward Ant for IAR Algorithm
Algorithm for Backward Ant of IAR Algorithm 44
1.11.7 Basic Ant Based Routing (BABR) for Wireless Sensor Networks (WSN) Ant Colony Optimization (ACO) [49, 50, 53, 72], is nature-inspired metaheuristic for solving complex Combinatorial Problems (CO). The main component of ACO algorithm is Pheromone Model. ACO, being an optimization approach, used to solve complex problems by iterating the following two steps: 1)
Using a Pheromone model, that is, a parametrized probability distribution over the solution space;
2)
The candidate solutions are used to modify the pheromone values in a way that is deemed to bias future sampling forward high quality solutions.
Basic Ant Based Routing leads to the development of AntNet Algorithm which can be summarized as follows: a)
Forward ant is launched from source node to sink node to determine the optimal path to destination.
b)
The main task of forward ant is to locate the food source with equal probability by using neighbouring nodes with minimum cost joining its source to sink.
c)
As ants move forward from node to node to reach the destination, the routing table gets updated side by side.
d)
Forward ants calculate all the information about the time length, congestion status and the node identifiers of the followed path.
e)
On reaching the destination node, the backward ant is created which follows the same path as forward ant, in opposite direction i.e. from food source to nest.
f)
During backward travel, local models of the network status and the local routing table of each visited node are modified by the agents as a function of the path they followed and of its goodness.
1.11.8 Ant Based Quality of Service Routing (ACO-QoSR) ACO-QoSR [76], a reactive routing algorithm was developed by Cai et. al in 2006 to tackle problems of constraint delay and energy in Wireless Sensor Networks. The 45
basic objective behind the development of ACO-QoSR routing is to find optimal routes between varied sensor nodes to sink node in such a way that the total end-toend delay is less than a boundary value, while the energy residual ratio i.e. ERR=Eresidual/Einitial is above a certain value. ACO-QoSR Algorithm When source node wants to send data, it first checks routing table to determine optimal path. Route probing will only start if there are no unexpired paths to the destination, and node needs to cache data waiting for transmission at the same time. Forwards ants does the task for route probing and after route discovery cached data is sent to destination in no time. In order to reduce time delay of route discovery, ACOQoSR algorithm starts a full route probe phase at the time of network initialization. Forward Ant Phase: In forward ant‟s phase, if the sending/source sensor unable to find a favourable path to sink node in routing table, it will generate a number of forward ants to search for optimal paths to destination. Forward ants will establish pheromone track between source to destination node. Forward ants comprise of various parameters: Timestamp origin, source and destination address. The main aim of forward ant is to collect intermediate node‟s local information and record the path information of various nodes from source to destination. Backward Ants Phase: When the forward ant reaches the destination, the forward ant will be killed and backward ant will be generated which carries source and destination address, backward ant ID, path information from forward ant and pheromone update value. Route Maintenance Phase: The entries in the routing table are basically pheromone values and probabilities that next-hop is a specific neighbour. Probabilities allow the ants to roam randomly in the environment and find new and optimal paths. Once the new optimal paths are discovered, the next hop probabilities are updated to routing table to reflect new paths from source nodes to sink nodes.
46
ACO-QoSR protocol was simulated in NS-2 Simulator considering the network of 100 sensor nodes in 1000x1000m area. ACO-QoSR protocol was compared with AODV and DSDV protocols on parameters like end-to-end delay, packet delivery ratio, routing overhead and path‟s normalized energy residual ratio. Simulation results showed that ACO-QoSR has better energy residual ratio and less overhead but packet delivery ratio is just average as compared to AODV and DSDV and routing overhead is small. 1.11.9 Ant Colony Optimization based Location-aware Routing (ACLR) Ant Colony Optimization based Location Aware Routing (ACLR) [77], a HighPerformance Routing Protocol for Wireless Sensor Networks was designed by Wang et. al in 2008. The principle behind the working of ACLR protocol is determination and selection of next hop by ants to a subset of the set of the neighbours of the current node which guarantee for the packet delivery rather than searching of whole neighbours to avoid loops. The protocol also determines the amount of pheromone which laid by the ant from source node to sink node. In addition, the protocol also proposes a novel scheme to evaporate the pheromone on the different segments of a certain route as per residual energy and the location information of nodes.
47
ACLR Algorithm ACLR Algorithm was simulated on OPNET Simulator and compared with 4 algorithms: Basic Ant Routing (BAR), SC, FP and IAR using network area of 200x300 m and 10000 sensors. Performance of ACLR algorithm is determined on energy consumption, efficiency and packet delivery latency. Results showed that ACLR consumes less energy as compared to 4 other algorithms and it is also better in terms of Packet Delivery.
1.12 Energy Efficient Routing Protocols based on Ant Colony Optimization for Wireless Sensor Networks In this section, various energy efficient routing protocols based on Ant Colony Optimization for Wireless Sensor Networks (WSN) are enlisted: 1.12.1 Ant Chain Protocol Ant-Chain, a hierarchical protocol is an energy efficient algorithm for Wireless Sensor Networks was being developed by Ding and Liu [78] in 2005 with primary focus towards energy efficiency, maximizing the lifetime of the sensor node and data integrity. It is basically a centralized algorithm in which the responsibilities of sensor nodes and the sink nodes are partitioned depending on their hardware resources and relative distances in order to optimize the energy and reduce transmission delays. AntChain algorithm is useful in those applications where the sensor nodes location is known well in advance. Ant Chain has an edge over other algorithms like LEACH and PEGASIS on the ground of energy efficiency. The drawback of this algorithm is that it is centralized in nature which in turn eliminates the capability of robustness in it. 1.12.2 Ant Aggregation The development of Ant Aggregation [79] was done by Misra and Mandal in 2006 on the grounds of argument that a multi hop communication model coupled with innetwork aggregation can lead to the reduction in consumption of energy by sensor
48
nodes and which in turn leads to the network lifetime enhancement. The protocol being developed targets the problem of optimal aggregation in multicast tree which is basically a NP-hard problem. It is a hierarchical protocol. The main objective of this algorithm was to build minimum cost aggregation trees, under which forward ants either look for path which is shortest to the sink note or for a close by aggregation points. In every node, a forward ant is unicast to the next hop with a certain probability defined in the protocol. So, we can say that Ant Aggregation performs better in terms of energy efficiency when applied to the sensor network. 1.12.3 Pheromone Based Energy Aware Directed Diffusion (PEADD) Pheromone Based Energy Aware Directed Diffusion (PEADD) protocol [80] is a data centric protocol and is regarded as another variant of Directed Diffusion and is based on Ant Colony Optimization heuristic. The protocol was developed by Zhu in 2007 to enhance the lifetime of sensor networks by only involving high energy nodes in the process of data gathering. In this algorithm ants increase the pheromone on a path proportionally to the remaining energy levels of the node. As a result, the paths with larger residual energy are increased and others are reduced. The level of pheromone is updated keeping in mind the amount of transmitting data. PEADD utilizes the general ant based routing in terms of selection and updation of route. 1.12.4 Ant Colony Multicast Trees (ACMT) Ant Colony Multicast Tress (ACMT) algorithm was proposed by De-min et al in 2008 [81] with the purpose to enhance the lifetime of sensor network by taking energy consumption into consideration. In ACMT algorithm, ants locate the trees with all destination nodes. Every node on the tree which has been found is current node. Every step made by each ant has no other meaning of any path than to enable the current tree to grow further. This algorithm returns positive feedback mechanism of basic ant colony algorithm. The developers of this algorithm have compared this with YANG model and Flooding and have observed significant improvements with regard to performance via simulation. 49
1.12.5 Improvised Ant Colony Routing (IACR) Improvised Ant Colony Routing (IACR), a hierarchical protocol, was proposed by Peng et al in 2008 [21] which is basically an improvement over ant colony routing protocol. The main advantage of this protocol that it takes into consideration the utilization of energy by sensor nodes along with QoS (Quality of Service) parameter. The algorithm being proposed consists of two parts: 1)
Routing Discovery- It is done in the same way as in basic ant colony routing protocol.
2)
Route Maintenance- The routing table is updated with the changes at the same time when the topology changes.
The researcher-Peng et al. has compared the proposed algorithm i.e. IACR with DD (Directed Diffusion) protocol using OmNet++ simulator and has shown a radical change with respect to energy conservation along with improvement to the problem of packet delay. 1.12.6 ACO Router Chip In order to optimize routing paths based on ACO algorithm and to attain reliable data communication in the network especially in the case of node failure, Okdem and Karaboga [82] in 2009 have developed an algorithm using ACO router chip. The main purpose behind the development of this algorithm was to achieve maximum lifetime and efficient data transmission. The researcher utilized event based simulator and the results have proved that the proposed algorithm performs better with regard to energy efficiency as compared to EEABR protocol. In addition to this, the researcher has also tested the proposed algorithm on ACO router chip. The Algorithm for the working of ACO Router Chip is as follows: 1)
The nodes which wants to transmit fragments the packets into parts and transmits the data on different paths. The packet when reaches the destination sends an acknowledgement message which shows its receipt. If there is noacknowledgement, the node retransmits the message. 50
2)
The sensor nodes which are having good energy level are chosen for data transmission which in turn results in the increase in network lifetime.
3)
The routing mechanism which lays the foundation for transmission of data from source node to sink node is based on the behavior of ants.
1.12.7 Energy Balanced Ant Based Routing Protocol (EBAB) Another location-based protocol for achieving energy efficiency and enhancing lifetime was proposed by Wang et al in 2009a [85]. In this protocol, the algorithm proposed was sub-divided into: Inter-Cluster and IntraCluster. In Intra-Cluster the clusters are created at the start of the routing process. The intracluster comprises the completion for cluster heads where cluster heads competed for based on the areas of which each node belongs because of the base station strength. As soon as the cluster selection is over, the cluster is finally set-up and the cluster head informs via message to all the nodes in the respective area of being cluster head. In turn all the nodes send „ACK‟ message confirming for joining themselves in the respective cluster. If any node in the network receives the join message from more than one cluster head, the node will choose the cluster head on two parameters: Distance between node and cluster head and energy required for transmission. During data transmission, all the nodes starts the receiver and it is the duty of the cluster head to inform all the nodes regarding TDMA time slot information which is the foundation and the only parameter which manages the data transmission. Every node has to reserve time slot for itself for transmission. The Intra-cluster makes use of ACO algorithm for transmission of packets in the network. The researchers have found that EBAB protocol performs much better in terms of energy management and packet delivery ratio and is much efficient as compared to existing protocols in terms of routing management and cluster formation also.
51
1.12.8 Adaptive Clustering for Energy Efficient WSN based on ACO (ACO-C) ACO-C is a location-based, another adaptive clustering energy efficient routing protocol was implemented by Ziyadi et al in 2009 [84]. The protocol basically works on two parameters: Distance minimization between the source node and sink node; Data Aggregation among all the sensor nodes. The researcher has simulated the proposed algorithm in MATLAB and has also compared the results with LEACH. It was being observed after simulation that ACO-C performs better in terms of energy efficiency but results are not impressive as compared to LEACH-C and PSO-C protocols. 1.12.9 Ant Colony Clustering Algorithm (ACALEACH) ACALEACH (Location Protocol) is a clustering algorithm and is regarded as enhanced version of LEACH Protocol being developed by Wang et al [85] in 2009b. During the selection of cluster head, the routing algorithm takes into consideration the remaining energy level of the sensor node and also the distance between the two cluster heads. In addition to this, it makes use of inter-cluster ACO algorithm to reduce the energy overhead of cluster heads. The algorithm has shown significant improvement in energy efficiency and has provided prolonged lifetime to the sensor network as compared to LEACH in the MATLAB simulation environment. 1.12.10 Ant Colony Optimization based- Energy-Aware Multipath Routing Algorithm (ACO-EAMRA) Ant Colony Optimization, a hierarchical protocol based-Energy Aware Multipath Routing Algorithm (ACO-EAMRA) was being proposed by Xia and Wu in 2009 [23]. The motivation behind the development of this algorithm is to conserve the energy to be utilized by sensor nodes during transmission. For this, the algorithm takes into consideration: the power level which is available with the nodes and energy which is being consumed by sensor nodes during route selection. The algorithm which is being proposed performs well in terms of energy efficiency as compared to DD algorithm.
52
1.12.11 Energy Efficient ACO Based QoS Routing (EAQR) This is a Hierarchical protocol and was developed by taking into consideration the deployment of sensors in the sensor networks where there is non-uniformity in traffic and in turn various problems like congestion which hampers the performance of network. So, Jietai et al. in year 2009 [86] proposed Energy Efficient ACO based QoS routing protocol based on Ant Colony Optimization. The main parameters on which this algorithm proposed is concentrated is QoS and balancing energy consumption in sensor network. By introducing metrics like minimum path energy and path hop count and by means of advancing pheromone trail model of the ant colony system, the algorithm provides various ways to improve performance of real time and common traffic. The algorithm being proposed by researchers was successfully simulated in NS-2 simulator and has shown performance improvements in terms of Average ETE delay, Average ETE delay jitter, Packet delivery ratio. 1.12.12 Comprehensive Routing Protocol (CRP) This is Data-Centric Protocol. Comprehensive Routing Protocol was developed by Guo et al in 2010 [87]. It is based on ant colony optimization and is regarded as improvised version of Energy Aware Routing (EAR). CRP protocol in its routing methodology uses probability of selection and improves network lifetime and packets arrival time. CRP protocol lays the foundation on the fact that always using the path which is considered as best and optimal path is not always the best as it can lead to the loss of path nodes energy and the protocol proposes sub-optimal paths. CRP protocol has three phases:
Routing Table Setup
Data Communication, and
Route Maintenance
53
1.13 Organization of Thesis The thesis is organized as follows: Chapter 1 provides information with regard to Wireless Sensor NetworksIntroduction, Design Principles and Challenges, WSN types, WSN Standards, Classifications of WSN network, Overview of sensor network routing and discussion with regard to basic WSN routing protocols along with routing protocols performance comparison. The chapter outlines Swarm Intelligence, Ant Colony Optimization and various Ant-Inspired Routing Protocols for Wireless Sensor Networks along with performance comparison of routing protocols for WSN based on Ant Colony Optimization. Chapter 2 provides comprehensive literature review acting as foundation and stepping stone of research being carried out for development of Improvised Energy Efficient ACO based Multipath routing protocol for sensor networks. Chapter 3 states Research Methodology and highlights Motivation behind Research, Research Methodology undertaken to perform research and enlists research gaps. Chapter 4 provides an overview of proposed energy efficient routing protocol: Improvised Energy Efficient Multipath ACO based Routing Protocol (IEEMARP)Design Parameters, Algorithm, Protocol Operation and Architecture highlighting the in-depth working of proposed protocol. Chapter 5 compares the performance of the proposed protocol via Simulation tool based experiments and compares with other ACO based routing protocols on different performance metrics.
Summary Wireless Sensor Networks, have wide range of exciting applications in virtually all fields of science, engineering like industry, military, security, environment, agriculture, health care and many more. Researchers face lots of critical challenges that needs to be sorted before these applications become reality. Furthermore, WSN
54
must deliver the data of interests via proper routing protocols considering various parameters like Energy efficiency, throughput, routing overhead, end-to-end delay and packet delivery ratio. Swarm Intelligence has come to rescue the problems of real world where for normal human beings it is highly difficult to understand and interpret. Swarm Intelligence based techniques like Ant Colony Optimization offers an alternative way of proposing and designing “Intelligent Systems” in which autonomy, emergence and distributed functioning replace control, pre-programming and centralization. Ant Colony Optimization (ACO) has roots in various current engineering fields especially WSN to solve various issues of routing complexity and propose energy efficient routing protocols to enhance network lifetime.
55
Chapter – 2
Literature Review This chapter outlines Literature Review about the problem undertaken in the research. Literature review covered in this chapter enlists various research papers with regard of review of sensor networks, routing protocols of sensor networks, energy efficient routing protocols of sensor networks, swarm intelligence, Ant Colony Optimization (ACO) and Energy Efficient routing protocols for sensor networks based on ACO. I.F. Akyildiz et al., Wireless Sensor Networks: A Survey [2] outlined the concept of wireless sensor networks regarding design and sensing factors and also stated varied algorithms and protocols for each layer in sensor based communication network. The paper highlights various open research issues for the realization of sensor networks. Yick et al., Wireless Sensor network survey [3] stated comprehensive literature review regarding WSN and highlighted new applications and issues of sensor communications. The paper outlined survey on three different categories: internal platform and underlying operating system, WSN protocol stack and network services, provisioning and deployment issues. The paper also presented various developments and challenges for researchers in area of sensor communications. Chong, C. Y., & Kumar, S. P. (2003). Sensor networks: evolution, opportunities, and challenges
[4]
provided
comprehensive
coverage
on
history
of
sensor
communications especially developments done by DARPA, DSN and SensIT. The paper outlined various research results in different sensor algorithms- network, localization and directed diffusion. The paper stated various recent trends in development of WSN and technical challenges surrounding deployment of WSN. Houshyarifar, V., & Amirani, M. C. (2014). Wireless Sensor Networks-A Review [88] outlines major challenges of WSN with regard to network layers and data flow. The paper states various approaches for efficient routing, package management and node localization for sensor networks. 56
Baronti et al. Wireless Sensor Networks: A Survey on the state of the art and 802.15.4 and ZigBee Standards [89] highlighted standards of Sensor networks, 802.15.4 and ZigBee. The paper also presented latest progress on state of art research on energy efficiency, networking, data management and security and also proposed some solutions to common issues in standards. Arampatzis, T., Lygeros, J., & Manesis, S. (2005, June). A survey of applications of wireless sensors and wireless sensor networks [90] outlined research issues and solutions to common problems of sensor networks for wide adaptability in diverse applications. The paper also surveyed various real time applications of sensor networks and researcher proposed important parameter titled “Mobility” for improvising the overall quality of sensor networks. Rashid, B., & Rehmani, M. H. (2016). Applications of wireless sensor networks for urban areas: a survey [91] highlighted the importance of deployment of WSN in urban areas and also focussed on different issues surrounding the technical viability of deployment of WSN. The paper enlisted various solutions to problems for efficient deployment of sensor networks in urban areas. Yuan, D., Kanhere, S. S., & Hollick, M. (2016). Instrumenting Wireless Sensor Networks—A survey on the metrics that matter [92] presented a survey on various important metrics to characterize the performance of sensor networks. The paper proposed system model for sensor networks and presented categories of various metrics that lays the foundation for determination of performance of sensor networks. The metrics highlighted include various real time practical aspects for live implementation of sensor communications. Rawat, P., et al. (2014). Wireless sensor networks: a survey on recent developments and potential synergies. [93] outlined recent developments in sensor network technologies and also highlighted research projects, standards, technologies and platforms for sensor based wireless networks. The paper stated potential of sensor networks in different applications and also highlighted the importance of sensor based applications. 57
Guo, W., & Zhang, W. (2014). A survey on intelligent routing protocols in wireless sensor networks. [94] discusses various routing protocols for wireless sensor networks based on reinforcement learning (RL), Ant colony optimization (ACO), fuzzy logics (FL), Genetic Algorithms (GA) and Neural networks (NNs) and further analysed the performance of routing protocols on basis of network lifetime efficiency. The author analyzed that ACO based routing protocols are best in improvising overall network lifetime of sensor nodes and ACO based routing protocols accommodate changes to changing environment in easy manner. Akkaya, K., & Younis, M. (2005). A survey on routing protocols for wireless sensor networks. Ad hoc networks [95] outlined a comprehensive survey of various routing protocols for sensor networks with proper classifications on basis of data centric, hierarchical and location based. The paper also highlighted various open research issues in area of routing especially energy efficiency, data aggregation etc. for researchers to work on. Goyal, D., & Tripathy, M. R. (2012, January). Routing protocols in wireless sensor networks: A survey [96] surveyed routing protocols for sensor networks and highlighted the pros and cons of each and every protocol for ensuring reliable and efficient multi-hop communication under varied scenarios. Al-Karaki, J. N., & Kamal, A. E. (2004). Routing techniques in wireless sensor networks: a survey [7] presented a comprehensive survey of routing techniques on the basis of multi-path, query-based, negotiation-based, QoS-based and coherent-based for wireless sensor networks. The research paper highlights performance comparison of routing protocols especially in energy saving parameter and also covers advantages and performance issues of each routing protocol for sensor networks. Vijayanand, S., & Suresh, R. M. (2007, December). An overlook on routing techniques in wireless sensor networks [14] outlines various design challenges for routing protocols in sensor networks and also listed comprehensive coverage of routing techniques i.e. Flat, Hierarchical and Location based routing. The researcher further proposed classification of these protocols on basis of multi-path, query, negotiation, QoS and coherent for effective protocol operation in real-time scenarios. 58
Pantazis, N. A., Nikolidakis, S. A., & Vergados, D. D. (2013). Energy-efficient routing protocols in wireless sensor networks: A survey [97] surveyed energy efficient routing protocols for sensor networks and listed the issues in each and every protocol for enhancing energy efficiency in overall network. Rault, T., Bouabdallah, A., & Challal, Y. (2014). Energy efficiency in wireless sensor networks: A top-down survey [98] presented top-down survey of trade-offs between application requirements and lifetime enhancement that arise when designing sensor networks. The researcher outlined a novel classification of energy saving techniques and also performed comprehensive survey on techniques to attain trade-offs between multiple requirements like multi-objective optimization. Sharma, P., & Kaur, I. (2015). A Comparative Study on Energy Efficient Routing Protocols in Wireless Sensor Networks [99], presented a comprehensive coverage of energy efficient routing protocols along with in-depth comparison of pros and cons of each and every protocol and also highlighted the various applicability scenarios of protocols to lay strong foundation for researchers to choose the best routing protocol for saving lifetime and attaining overall efficiency in sensor environments. Merkle,
D.,
&
Middendorf,
M.
(2014).
Swarm
intelligence.
In Search
methodologies [29] outlined in comprehensive manner the concept of Swarm Intelligence in the field of computer science to solve complex computational problems via ACO and PSO metaheuristics. The paper highlighted the applicability of swarm intelligence algorithms along with other metaheuristics to solve various issues. The concept of ACO and PSO is highlighted in comprehensive manner in this paper. Dressler, F., & Akan, O. B. (2010). A survey on bio-inspired networking [31] demonstrates the concept of bio-inspired networking and highlighted some of the open research issues for researchers. The paper further highlights the concept of nanonetworks and also highlights some research projects and funding agencies for taking up Bio-inspired networking to next level of implementation. Garnier, S., Gautrais, J., & Theraulaz, G. (2007). The biological principles of swarm intelligence [24] emphasized on the role of interactions and importance of 59
bifurcations that appear in the collective output of the colony when parameters change. The researcher categorized the collective behavior displayed by insects according to the functions that emerge at the level of colony and organize its global bahavior. The researcher also stated the role of modulations of individual behaviors by disturbances in overall flexibility of insect colonies. Chu, S. C., Huang, H. C., Roddick, J., & Pan, J. S. (2011). Overview of algorithms for swarm intelligence [100] highlighted various algorithms in the area of swarm intelligence that can be utilized for optimization. The researcher has outlined Particle Swarm Optimization (PSO), Ant Colony System (ACS) and Artificial Bee Colony (ABC) along with comparisons between the protocols. Fountas, C. (2010). Swarm Intelligence: The Ant Paradigm [101] stated the concepts of Swarm Intelligence based Ant colony optimization framework and also highlighted the three most popular ACO based algorithms- Ant System, Ant Colony System and MAX-MIN Ant System and also analysed the design changes between the algorithms. Kordon, A. K. (2010). Swarm intelligence: The benefits of swarms [102] outlined the crucial advantages of artificial swarms along with unique advantages of swarm intelligence. The researcher also highlighted thorough and comprehensive literature survey on swarm intelligence techniques- ACO and PSO and signified the advantages of implementing the techniques for solving complex computational problems in real world. Abraham, A., Guo, H., & Liu, H. (2006). Swarm intelligence: foundations, perspectives and applications [103] stated the design and implementation of ACO and PSO algorithms for finding solutions to various optimization problems, applications with regard to real world and data mining. The paper also highlights the results via implementing ACO based algorithms for data mining and clustering of web server. Blum, C., & Li, X. (2008). Swarm intelligence in optimization [28] highlighted the implementation of two Swarm Intelligence based algorithms- ACO and PSO and stated the origins of both algorithms and highlighted the examples of real time
60
applications of implementing ACO and PSO based techniques for solving complex computational problems. Parpinelli, R. S., & Lopes, H. S. (2011). New inspirations in swarm intelligence: a survey [104] highlights various social insects based swarms like Bacterial foraging, glow-worms, fireflies, slime moulds, cockroaches, mosquitoes and other organisms and stated the respective behavior for usage by varied researchers in solving problems via biological behaviors and optimizations. The paper also highlighted the most recent and important real time applications along with features of meta-heuristics. Dorigo, M., Birattari, M., & Stutzle, T. (2006). Ant colony optimization [46] proposed Ant Colony Optimization- A Computational Swarm Intelligence technique based on real ants. The researchers have proposed a novel algorithm “Ant System” to be applied to Travelling Salesman problem for obtaining optimized results. The researchers have highlighted the importance and applications of ACO applicability in varied applications. Blum, C. (2005). Ant colony optimization: Introduction and recent trends [105] outlined various algorithms based on Ant Colony System and also provided comprehensive coverage of implementing ACO in context of discrete optimization and how ACO can be applied to solve complex optimization problems. The paper also highlights various research directions for applying ACO in various applications from artificial intelligence to future oriented areas of computer science. Mohan, B. C., & Baskaran, R. (2012). A survey: Ant Colony Optimization based recent research and implementation on several engineering domain [106] outlined the in-depth coverage of Ant Colony Optimization and review on latest research and implementation of ACO. The researcher has proposed a novel ACO based model for solving network routing problem and compared the model with traditional routing algorithms. The results of ACO as compared to other traditional algorithms are much better and author concluded that ACO based network routing model is better in terms of efficiency and optimal routing.
61
Blum, C. (2007, September). Ant colony optimization: introduction and hybridizations [107] reviewed Ant Colony Optimization and presented the optimized mathematical model of ACO with regard to solution construction, pheromone update and also stated the ACO hybridizations and real time application area like TSP, Quadratic assignment problem, scheduling, timetable, Cell Placement etc. Pei, Y., Wang, W., & Zhang, S. (2012, March). Basic ant colony optimization [108] proposed an Improved ACO for multicore computing. The modified ACO algorithm is being proposed for solving complex computation tasks of multicore computing via mathematical models and theoretical proofs. Shah, S., Bhaya, A., Kothari, R., & Chandra, S. (2013). Ants find the shortest path: a mathematical proof [109] proposed a new modified ACO algorithm called “EigenAnt” for determining the shortest path from source to destination based on pheromone removal that exists on the path which is choosen for each trip. EigenAnt algorithm is tested and compared to basic ACO via simulation and it is observed that algorithm is more successful when applied to complex combinatorial optimization problems. Monteiro, M., Fontes, D. B., & Fontes, F. A. (2012). Ant Colony Optimization: a literature survey [110] stated the comprehensive literature work for developing ACO based algorithms via varied approaches cited in literature in this paper. The researcher also stated that ACO based algorithms perform well in many combinatorial problems as compared to Genetic algorithms. Shtovba, S. D. (2005). Ant algorithms: theory and applications [111] outlined the theory and applications of ant algorithms based on simulation of self-organized colony of biological ants. The Researcher stated that ACO based algorithms successfully applied to travelling salesman, vehicle routing, graph coloring, job scheduling and network traffic optimization. The results concluded that Ant based algorithms are efficient for telecommunications network routing problems. Cordón García, O., Herrera Triguero, F., & Stützle, T. (2002). A review on the ant colony optimization metaheuristic: Basis, models and new trends [112] outlined 62
algorithms based on ACO and analysed the relationship between ACO and other heuristics. The paper also demonstrated some latest theoretical developments in ACO area via new fields and research directions. Dorigo, M. (2001, September). Ant algorithms solve difficult optimization problems. [113] signifies a general overview of Ant Colony Optimization along with its algorithms and highlighted the results on theoretical properties of algorithms based on ACO to solve complex combinatorial problems. Sim, K. M., & Sun, W. H. (2003). Ant colony optimization for routing and loadbalancing: survey and new directions [114] outlined a comparison and critique of state of art approaches for mitigating stagnation, surveyed major research problems in applying ACO based approach for routing and load balancing and listed new research directions and open research issues. The paper also stated the performance comparison of ACO based approaches like AntNet [17] and other add on extensions like ASGA and SynthECA for applying ACO based approach for optimization and load balancing. Zhang, G., Pérez-Jiménez, M. J., & Gheorghe, M. (2017). Fundamentals of Evolutionary Computation [115] stated comprehensive review of algorithms and technique
concerning
Genetic
Algorithms,
Quantum
Inspired
Evolutionary
Algorithms, Ant Colony Optimization, Particle Swarm Optimization and Differential evolution. The paper also stated the comprehensive comparison on basis of mathematical equations and proofs for solving complex computation problems. Yi, G., Jin, M., & Zhou, Z. (2010, June). Research on a novel ant colony optimization algorithm [116] proposed a novel ACO based algorithm based on adaptively adjusting pheromone decay parameter. The algorithm is applied via simulation to Travelling Salesman problem and results showed that proposed ACO algorithm is better in determining optimal routes as compared to basic ant colony routing algorithm. Zungeru, A. M., Ang, L. M., & Seng, K. P. (2012). Classical and swarm intelligence based routing protocols for wireless sensor networks: A survey and comparison [67] surveyed the routing protocols based on Swarm Intelligence for sensor networks and 63
thorough comparsion is presented on basis of computational complexity, network structure, energy efficiency and path establishment. The researcher also presented a powerful comparison of routing protocols presented in survey with different performance parameters via simulation based results using Matlab based simulatorrouting modeling application simulation environment (RMASE) and simulation results presented the pros and cons of each routing protocol. Saleem, M., Di Caro, G. A., & Farooq, M. (2011). Swarm intelligence based routing protocol for wireless sensor networks: Survey and future directions [68] outlined comprehensive survey of routing protocols based on swarm intelligence for wireless sensor networks based on Ant Colony and Bee Colony optimization. The researcher presented a taxonomy of routing protocols for efficient classification and also presented some key issues for future directions. Zengin, A., & Tuncel, S. (2010). A survey on swarm intelligence based routing protocols in wireless sensor networks [65] highlighted routing protocols for sensor networks based on swarm intelligence and also presented a comparison of protocols on basis of battery life, scalability, maintability, survivability, adaptability and other approaches. The researcher concluded that ACO based routing protocols are best for overcoming all sorts of routing issues and enhancing network lifetime. Zhang, Y., Kuhn, L. D., & Fromherz, M. P. (2004, September). Ant colony-based energy-aware multipath routing algorithm for wireless sensor networks [18] proposed three new routing algorithms: Sensor Driven Cost Aware Ant Routing (SC), Flooded Forward Ant Routing (FF), Flooded Piggybacked Ant Routing for sensor networks. The researcher has simulated the three proposed algorithms in Matlab based simulator-Prowler using network model for sensor networks and compared with existing basic Ant Colony approaches. The researcher concluded that proposed algorithms- SC in terms of energy efficiency, FF in terms of end-to-end delay and FP in terms of network optimality work better as compared to existing ant routing protocols and even traditional routing protocols like AODV.
64
Domínguez-Medina, C., & Cruz-Cortés, N. (2010). Routing algorithms for wireless sensor networks using ant colony optimization [118] presented the performance comparison of two Ant based routing protocols Ant Colony Optimization-Based Location-Aware routing for WSN (ACLR) and Energy Efficient Ant Based Routing Algorithm (EEABR). Simulations being performed on basis of Latency and Energy Consumption and results stated that EEABR is highly energy efficient as compared to ACLR and ACLR is better in end-to-end delay as compared to EEABR. The researcher concluded that overall EEABR performs better as compared to ACLR. Shirkande, S. D., & Vatti, R. A. (2013, April). Aco based routing algorithms for adhoc network (wsn, manets): A survey [119] outlined comprehensive review of ant based routing protocols for sensor networks (WSN) and Mobile Adhoc Network (MANETS). The paper also presented a performance comparison of algorithms on performance metrics, pheromone function to select proceeding node, simulator used for testing and energy efficiency. Misra et al. An ant swarm-inspired energy-aware routing protocol for wireless ad-hoc networks [120] proposed a novel energy efficient routing protocol based on Ant Colony optimization i.e. EAAR (Energy Aware Ant Based Routing) and performed hard core testing of EAAR using simulation. The paper concluded that simulation based results proved that EAAR is better as compared to AODV, AntHocNet and MMBCR in terms of mobility and energy efficiency. The researchers proved that EAAR is better in energy efficiency due to multipath routing, Ant-based path-search for energy abundant routes and efficient route and recovery mechanism. Kumar, N. A., & Thomas, A. (2012, July). Energy efficiency and network lifetime maximization in wireless sensor networks using improved ant colony optimization [121] proposed a novel data collection scheme, Maximum Amount Shortest Path (MASP) to improvise throughput and energy efficiency in sensor networks. MASP is tested via simulation based experiments using NS-2. The results concluded that MASP outperforms in terms of delay, bandwidth and energy efficiency in sensor networks. 65
Cheng, D., Xun, Y., Zhou, T., & Li, W. (2011). An energy aware ant colony algorithm for the routing of wireless sensor networks [122] proposed Energy Aware Ant Colony Algorithm (EAACA) for wireless sensor networks and compared the performance of EAACA on basis of theoretical model and simulation with ACA and researchers concluded that as compared to ACA, EAACA performs better in terms of maintaining overall energy efficiency in nodes and extending overall lifetime of sensor networks. Wang G., Wang Y, Tao x. An Ant Colony clustering routing algorithm for wireless sensor networks [83] proposed Ant Colony Algorithm (ACA) for inter-cluster routing to select the most efficient path of packet forwarding from cluster head to base stations. Simulation results stated that ACA is 30% efficient in enhancing overall network lifetime of sensor network as compared to LEACH protocol. Arya, R., & Sharma, S. C. Energy optimization of energy aware routing protocol and bandwidth assessment for wireless sensor network [123] outlined and compared the performance analysis of ACO based energy aware routing with without ACO energy aware routing and proved that ACO based energy aware routing algorithms are better in terms of optimal path routing and maintains overall energy efficiency in sensor networks. Okafor, F. O., & Fagbohunmi, G. S. (2013). Energy Efficient Routing in Wireless Sensor Networks based on Ant Colony Optimization [58] proposed heuristic method for conserving energy efficiency in sensor networks. The paper also compared three routing algorithms: Ant System, Ant Colony System and Improved Ant System in terms of sensor network routing. The results concluded that Ant Colony Optimization is better as compared to others in maintaining overall energy efficiency in routing and also enhances overall lifetime of network. Zungeru, A. M., Seng, K. P., Ang, L. M., & Chong Chia, W. (2013). Energy efficiency performance improvements for ant-based routing algorithm in wireless sensor networks [124] proposed Improvised Energy Efficient ant based routing (IEEABR) with enhanced functionalities in terms of intelligent initialization of 66
routing tables by giving utmost priority to neighbouring nodes, intelligent updation of routing table in case of any sort of link failure and maintaining congestion control by reducing flooding of ants. The results are compared with EEABR routing protocol via RMASE simulator. The researcher concluded that IEEABR is better in energy efficiency and suitable for dynamic environments as compared to EEABR and even SC and Bee sensor based routing protocols. Wang, L., Sun, Q., & Ma, H. (2010, March). Energy consumption optimize based on ant colony algorithm for wireless sensor networks [125] proposed a new routing protocol based on Ant Colony Algorithm (ACA) to optimize energy efficiency. The researcher concluded that proposed routing algorithm outperforms DD and basic ACO in terms of routing overhead, energy efficiency and overall network effectiveness. Xia, S., Wu, S., & Ni, J. (2009, December). A new energy-efficient routing algorithm based on ant colony system for wireless sensor networks [126] proposed a novel multipath algorithm based on ant colony system with enhancements based on Ant Marginalization rule, State transition rule and Global Pheromone update rule to provide effective solutions to local convergence, optimization and multipath. The proposed algorithm is tested using simulation and compared to other Ant colony algorithms and results provided that algorithm has significant edge over energy efficiency and overall network lifetime in sensor networks. Almshreqi, A. M. S., Ali, B. M., Rasid, M. F. A., Ismail, A., & Varahram, P. (2012, February). An improved routing mechanism using bio-inspired for energy balancing in wireless sensor networks [127] proposed a self-optimization scheme for sensor networks based on ACO heuristics and method is highly effective in distributing traffic load in network to extend energy and overall lifetime. The protocol is tested using simulator and results proved that proposed protocol is better as compared to EEABR protocol in energy efficiency. Jiang, X., & Hong, B. (2010, June). ACO based energy-balance routing algorithm for WSNs. [128] proposed ACO based Energy balance routing algorithm (ABEBR) and introduced a pheromone update operator to improvise the overall energy efficiency 67
and via simulation ABEBR is compared to LEACH, DD and Flooding protocols. The researcher concluded that ABEBR is better in overall energy efficiency and provides enhanced network lifetime in sensor networks. Orojloo, H., Moghadam, R. A., & Haghighat, A. T. (2012). Energy and path aware ant colony optimization based routing algorithm for wireless sensor networks. [129] proposed Energy and Path aware Ant Colony Optimization based routing protocol (EPWSN) for sensor networks via pheromone update operator to efficiently integrate energy consumption, hop count and path length in routing protocol. EPWSN is hard core tested via simulation. Simulation results proved that EPWSN is better as compared to ACLR algorithm. The paper concluded that proposed algorithm EPWSN is better in Energy Efficiency, reducing routing overhead and better packet delivery ration as compared to other ant based routing protocols. Lin, Y., Zhang, J., Chung, H. S. H., Ip, W. H., Li, Y., & Shi, Y. H. (2012). An ant colony optimization approach for maximizing the lifetime of heterogeneous wireless sensor networks [130] proposed ACO-MNCC protocol for optimizing energy efficiency of heterogeneous wireless sensor networks. The approach makes use of pheromone and heuristic information to search for optimal paths and provides the best solution. ACO-MNCC is tested via simulation and results state that the proposed methodology is better in providing path optimality, energy efficiency in sensor networks. Hui, X., Zhigang, Z., & Xueguang, Z. (2009, July). A novel routing protocol in wireless sensor networks based on ant colony optimization. [131] proposed novel ACO based routing protocol for enhancing node energy to achieve QoS in dynamic and adaptive routing to overall attain energy efficiency in WSN. Simulation results concluded that proposed ACO routing protocol is better in enhancing overall network lifetime. Arabshahi et al. Adaptive routing in wireless communication networks using swarm intelligence [132] outlined the importance of adaptive routing in wireless sensor networks. Researcher concluded via mathematical model and proofs, to enhance 68
overall efficiency in sensor communications via adaptive routing, swarm intelligence approaches are best in power efficiency and overall scalability and robustness. Laxmi, V., Jain, L., & Gaur, M. S. (2006, December). Ant colony optimization based routing on NS-2 [133] outlined thorough comparison of Ant Based Routing protocols like Ant Based Control Routing, AntHocNet and AntNet using NS-2 simulator on various performance metrics. The paper concluded that results show that Ant Based Routing techniques are better in maintaining overall efficiency in routing for sensor communications networks. Ali, Z., & Shahzad, W. (2013). Analysis of routing protocols in ad hoc and sensor wireless networks based on swarm intelligence [134] outlined the detailed comparison of swam intelligence based routing protocols for wireless sensor networks and MANETS. The researcher concluded that ACO and PSO based routing protocols are highly suitable for routing efficiency in sensor communications. Ahmed, M. B., Boudhir, A. A., & Bouhorma, M. (2012). New routing algorithm based on ACO approach for lifetime optimization in wireless sensor networks [135] propose novel routing protocol based on Ant Colony Optimization for Wireless Sensor Networks. The proposed protocol is compared with traditional routing protocols like AODV and LEACH. Simulation results stated that ACO based routing protocol proposed is better in terms of energy efficiency as compared to other routing protocols. Sun et al. (2010) A Routing Protocol based on Ant Colony Algorithm for Wireless Sensor Networks [136] proposed Directed Diffusion based Ant Colony Algorithm (DDBA) based on Ant Colony Algorithm for optimizing energy efficiency in routing. DDBA is compared to Directed Diffusion (DD) protocol via theoretical mathematical model and simulation. The results stated that DDBA is much efficient in balancing load, energy utilization as compared to DD and is highly suitable for sensor networks. Xia, S., & Wu, S. (2009, November). Ant colony-based energy-aware multipath routing algorithm for wireless sensor networks [23] proposed multipath based energy efficient routing protocol for sensor networks for optimizing energy efficiency of 69
nodes and maintaining route optimality in overall network. The algorithm proposed is compared to DD protocol and via simulation the result stated that proposed algorithm is better in self-adaptability, energy efficiency and mobility to improvise overall accuracy and scalability in sensor networks. Norouzi, A., Babamir, F. S., & Zaim, A. H. (2011). A novel energy efficient routing protocol in wireless sensor networks [137] proposed a novel protocol Fair Efficient Location based Gossiping (FELGossiping) to overcome gossiping issues in sensor communications. The protocol is tested via simulation and compared with ElGossiping, LGossiping and Gossiping on parameters like Packet Delivery, Energy and Delay. The paper concluded that FELGossiping is much better in providing energy efficiency and maintaining overall bandwidth in network and in addition the protocol also reduces packet loss which makes the sensor network overall optimal and efficient network. Ahvar, S., & Mahdavi, M. (2011). EEQR: An energy efficient query-based routing protocol for wireless sensor networks [138] proposed Energy Efficient Query-based Routing Protocol (EEQR) to optimize energy level of nodes in network. EEQR protocol is tested and compared with Drumor, Flooding and Rumor based protocols on energy consumption. The paper concluded that EEQR is better and improvised network lifetime as compared to other query-based protocols in sensor networks almost more than 18 to 25%. Gajjar, S., Sarkar, M., & Dasgupta, K. (2014). Self-Organized, Flexible, Latency and Energy Efficient Protocol for Wireless Sensor Networks [139] proposed SelfOrganized, Flexible, Latency and Energy Efficient (SOFLEE) protocol for wireless sensor networks based on TDMA combined with Multihop routing information. The protocol is compared with FlexiTP, SOTP, EEFF and D-MAC protocols via simulation on metrics like energy efficiency, routing performance and link efficiency. The paper concluded that SOFLEE is overall the best protocol in attaining energy efficiency as compared to other routing protocols for sensor networks. Lin, T. L., Chen, Y. S., & Chang, H. Y. (2014, August). Performance Evaluations of an Ant Colony Optimization Routing Algorithm for Wireless Sensor Networks. [140] 70
compared ACO routing protocol with AODV, DSR AND DSDV using NS-2 simulation environment. The paper concluded that ACO based routing protocol outperforms AODV, DSR and DSDV in packet delivery ratio, energy efficiency but performance lags in end-to-end delay. Guo, H. (2012, June). Investigation on ant-colony based routing algorithm for wireless sensor networks [141] proposed IACA (Improvised Ant Colony Algorithm) by improvising shortest path determination via restricted pitch point energy consumption. The IACA is compared with ACA via mathematical formulas and proof and results outline that IACA is better in terms of link establishment, path discovery, hierarchical routing and overall energy efficiency for sensor networks. Tong, M et al. (2015) An energy-efficient multipath routing algorithm based on ant colony optimization for wireless sensor networks [142] proposed Energy Efficient ACO-Based Multipath Routing Algorithm (EAMR) for effective path discovery and route maintenance. The protocol proposed has improvisations in terms of ant packet structures, pheromone update formula, pheromone update mode and multipath mechanism scenario. EAMR protocol is compared with AOMDV, EEABR, EAMR and AntHocNet on parameters like End to end delay, energy efficiency and Routing Overhead via simulation. The paper concluded that EAMR is much better in terms of packet overhead delay removal, energy consumption and routing efficiency as compared to other protocols. Saleem, K et al. (2009). Ant based self-organized routing protocol for wireless sensor networks [143] proposed enhanced routing protocol based on self-organizing ant colony algorithm to improvise end-to-end delay, routing overhead and energy efficiency in sensor networks. The algorithm is improvised in terms of permanent loop avoidance to combat dead lock issue in wireless networks. Proposed protocol is tested and implemented using NS-2 Simulator and compared to existing ACO algorithms. The paper concluded that proposed algorithm is improvised in energy and routing efficiency. 71
Wang, Z., & Zhang, D. (2005, November). An improved ACO algorithm for multicast routing [144] proposed improvised ACO algorithm via local and global pheromone updating rule and to improvise efficiency and multipath routing in WSN. The paper concluded that experimental results prove the proposed routing algorithm better in optimal path search, scalability, robustness and overall efficiency in wireless sensor network multipath routing. Singh, A., & Behal, S. (2013). Ant colony optimization for improving network lifetime in wireless sensor networks [145] proposed Energy Efficient Shortest Path Routing Algorithm for improvising path discovery, link establishment and energy efficiency in sensor networks based on ACO multipath routing algorithms. The proposed algorithm is compared with Termite-Hill, AODV and FF on basis on path search, energy efficiency and network lifetime via simulation. The paper concluded that Energy Efficient Shortest Path Routing Algorithm provides better network lifetime and maintains good amount of energy efficiency of nodes. Li, Z., & Shi, Q. (2013). An QoS Algorithm Based on ACO for Wireless Sensor Network [146] proposed novel energy efficient QoS routing algorithm based on Ant Colony for improvising overall network lifetime of network. The proposed protocol enhances ACO algorithm via SNGF in optimizing routing tables, load balancing and energy efficiency. The paper concluded that simulation based results proved that proposed algorithm maintains overall energy efficiency and enhances network lifetime of sensor nodes. Shen et al. (2009) A routing algorithm based on ant-colony in wireless sensor networks [147] proposed ACO enhanced routing protocol based on LEACH protocol. The protocol proposed improvised the cluster head selection and also maintains the traffic load in overall network. The paper concluded that Simulation based results proved that proposed protocol is improvised in metrics like energy consumption, network lifetime and scalable routing in dynamic changing environments. Kate, V. B., & Das, S. (2014). Energy Efficient Ant Colony Optimization based Routing Protocol for Wireless Sensor Networks [148] proposed novel ACO based
72
routing protocol to combat the issue of energy efficiency by improvising routing decision on ACO pheromone parameter. The proposed protocol is hard core tested via NS-2 simulator and compared with traditional protocols as well as ACO based basic routing protocols. The paper concluded that proposed protocol is highly efficient in link discovery, path establishment, routing overhead and energy efficiency about 46% better as compared to other routing protocols. Ding, N., & Liu, P. X. (2005). A centralized approach to energy-efficient protocols for wireless sensor networks [78] proposed ACO based routing protocol for sensor networks “AntChain” to create near-optimal chain for develop efficient routing path for randomly created network of sensor nodes. AntChain protocol is tested via simulation results and compared with protocols like LEACH, PEGASIS. The paper concluded that AntChain protocol is much efficient in terms of energy efficiency, data integrity and maintaining overall network lifespan in sensor networks. Peng et al. An adaptive QoS and energy-aware routing algorithm for wireless sensor networks [21] proposed Improvised Ant Colony Routing (IACR) protocol to overcome various challenges like congestion, energy efficiency, computation of sensor nodes, packet dropping due to mobility. The protocol proposed is tested on OMNeT++ simulator and compared with Directed Diffusion (DD) protocol. The paper concluded that IACR is better in packet delivery ratio, load balancing, end to end delay and energy efficiency in sensor networks. Patel, M., Chandrasekaran, R., & Venkatesan, S. (2004, June). Efficient MinimumCost Bandwidth-Constrained Routing in Wireless Sensor Networks. [19] proposed EMCBR routing protocol for sensor networks for optimizing network lifetime and minimizing packet overhead ratio. The proposed protocol is compared with MLBCR via simulation. Simulation results concluded that EMCBR is much efficient in maintaining higher residual energy and routing efficiency in sensor networks. Vergados (2007) Enhanced route selection for energy efficiency in wireless sensor networks. [149] proposed Low Cost Min-Max Energy Routing (LCMMER) protocol for avoiding least-energy nodes during communication to maintain overall efficiency 73
in network. LCMMER is tested via simulation and compared with Minimum Total Transmission Power Routing (MTPR) and Min-Max Battery Cost Routing (MMBCR) over route selection strategies. The paper concluded that LCMMER is must efficient in determining optimal path and makes no use of low energy nodes and is highly sophisticated algorithm for maintaining overall energy efficiency in WSN. Sahni, J. P. S., & Park, J. (2005). Maximum Lifetime Routing in Wireless Sensor Networks [150] proposed shortest cost routing algorithm for optimizing energy efficiency and maintaining residual energy levels between two sensor nodes. Proposed algorithm was tested on NS-2 simulator and compared with other traditional routing protocols for WSN and results concluded that proposed algorithm is efficient in energy efficiency by solving linear programming problem. Sundani et al. (2011). Wireless sensor network simulators a survey and comparisons [151] outlined a comprehensive review of simulation tools available for simulating various concept of sensor networks. The author stated detailed comparison of simulation tools in different application environments along with pros and cons of each tool. Kellner, A., Behrends, K., & Hogrefe, D. (2010). Simulation environments for wireless sensor networks. [152] stated technical report concerning simulation tools available for sensor networks. Researcher considered diverse parameters like accuracy, performance, memory usage for determining and selecting the best simulator. The paper concluded with the comparison of simulation results with testbed results of sensor lab at telematics group. Egea-Lopez et al. Simulation tools for wireless sensor networks [153] discussed various simulation tools available for sensor networks simulation depending on applications and suitability. Researcher added a suitable model for WSN simulation for selecting the appropriate simulation tool along with comprehensive comparison of simulation tools available till date.
74
The above mentioned Literature Review has laid a strong foundation stone for the research being carried out to propose a novel routing energy efficient protocol based on Ant Colony Optimization for sensor networks. Till date, Wireless Sensor Networks are surrounded by critical issues like Packet Delivery Ratio, Optimal Latency, Security, Localization, Throughput, End to End delay and especially Energy Efficiency. The Literature survey presents a crystal-clear picture of the routing protocols being proposed till date for wireless sensor networks and enlists the shortcomings of each and every protocol proposed. With literature review, some of the protocols like ACEAMR, AntChain, EMCBR and IACR are shortlisted. Considering the shortcomings of these routing protocols in terms of Energy Efficiency, a novel routing protocol is proposed to maintain overall effectiveness in sensor network.
75
Chapter – 3
Research Methodology 3.1
Motivation
The main motivation behind this research is to investigate and propose some solution with regard to the problem of Energy Efficiency in Wireless Sensor Networks by considering Ant Colony Optimization- A Swarm Intelligence based technique. Ant Colony Optimization based algorithms are highly optimal in determining the shortest path from source to destination and demonstrate an inherent adaptability that can be utilized to find solutions to various dynamic problems especially optimal routing in a network. Swarm Intelligence based Ant Colony Optimization (ACO) technique is considered as one of the most promising area of research in various engineering fields. Ants move arbitrarily in the environment in search of food and demonstrate a spectacular ability to determine the shortest path from food source to nest. A single ant cannot determine the optimal path because of limited intelligence. So, other ants highly coordinate and cooperate among each other to find food, by laying a trail of chemical substance called “Pheromone”. The most important feature of ants is to adapt the dynamic nature of changing environment. Any occurrence of obstacle between food source and nest, ants are able to find the shortest path around the obstacle in such a matter that the path is again optimal from food source to nest. Various biologists have proved that many colony level behaviours observed in social agents called ants can be explained via simple models in which only stigmergic communication is present. These models are simulated via computer simulation software’s and it has been proved that Ant Based models are highly efficient in determining the network optimality. The remarkable success of social insects can be regarded as a starting point for new solutions to highly complex problems in area of Engineering especially computer science.
3.2 Research Problem Wireless Sensor Networks is regarded as the most important potential field of research interest in 21st century. Wireless Sensor Networks field is surrounded by 76
enhancement in Wireless communications, MEMS (Microelectronic mechanical systems) which has led to the development of low-cost, efficient power and smart sensors deployed in real time environments and networked using Wireless links and Internet to provide real time sensing data of varied applications like Industrial production, Military, Civilian, Health care, Agriculture and many more. Wireless sensor networks, being one of the most advanced and adaptable solution for real time sensing of data in mission critical and crucial applications but still is surrounded by lots of issues and challenges. The most important issues currently pertaining around WSN are Energy Efficiency, Optimal Routing, Security, Time Synchronization, Dynamic topology adaption and many more. Wireless Sensor Network routing is regarded as highly potential research field by researchers to propose energy efficient optimal path routing protocols via Swarm Intelligence. The overall quality of Wireless sensor network is determined by routing protocol which lay the foundation of fast transfer of data among nodes by visiting less number of nodes, in turn providing, Quality of Service (QoS), less packet delay and loss, less routing overhead, no congestion and overall maintaining energy efficiency in sensor nodes. As the energy level of nodes decline on transfer of packet it is highly important for routing protocol to take less number of hops and only wake up those nodes which are required for path between source and destination to make overall network up and ready for considerably longer period of time. Dorigo et al. have demonstrated an adaptive distributed, mobile agent based system, to solve complex problems of communication networks based on Ant Colony Optimization. Zhang et al. proposed three novel Ant based routing algorithms for wireless sensor networks- Sensor driven Cost-aware Ant Routing (SC) algorithm, Flooded Forward Ant Routing (FF) algorithm and Flooded Piggybacked Ant Routing (FP) algorithm and tested the three algorithms using Prowler simulator on parameters like Latency, Success rate, Energy consumption and energy efficiency and concluded that SC algorithm is highly energy efficient, FF has shorter delays and FP even having higher success rate but not energy efficient. SC, FP and FF are not suitable for optimal routing in Wireless sensor networks due to different limitations. 77
Patel et al. proposed Efficient Minimum-Cost Bandwidth-contained Routing (EMCBR) protocol to determining optimal route for data transfer between sensor nodes to base station node while guaranteeing the channel load not to exceed the overall capacity. The algorithm was compared with MLBCR protocol and simulation results that EMCBR is highly scalable and efficient for optimal routing in WSN. Ding et al. proposed AntChain algorithm to determine efficient path in random deployed WSN network. The algorithm was tested on NS-2 simulator with extended LEACH protocol version and results show that AntChain protocol provides highly energy efficient solution to solutions in those networks where base stations are nearby. Overall AntChain algorithm is regarded as highly flexible, reliable and efficient protocol to provide best QoS in data transfer. Peng et al. proposed Improved Ant Colony Routing (IACR) algorithm to provide novel solution to multi-constrained routing issue in sensor networks. IACR lays strong foundation for WSN optimal routing in terms of consuming less energy, low processing power and less memory storage. Simulation results proved that IACR algorithm is overall the best algorithm in varied WSN scenarios. Camilo et al. proposed Energy Efficient Ant Based Routing Algorithm for Wireless sensor networks (EEABR) to improvise routing paths. EEABR makes use of lightweight ants to determine optimal distance between source node to sink node considering hop count and energy level of nodes. Xia and Wu proposed Ant Colony Optimization based Energy Aware Multipath Routing Algorithm for wireless sensor networks (ACEAMRA). ACEAMRA algorithm considers exiting energy level of nodes and energy consumption of every path for selection of route, makes use of self-organization and self-adaptability and makes use of ACO based routing algorithm to determine the optimal path of packet routing from source nodes to base nodes. ACEAMRA maintains energy level of nodes by selecting the optimal path. Simulation based results proved that ACEAMRA is better in terms of load balancing and maintaining energy level of nodes but still faces lots of issues in terms of multipath routing. 78
Considering physical boundaries of sensor nodes and type of network topology used, various sensor networks makes use of multi-hop routing algorithms to improvise the overall network life time. Various real time applications like Battlefield monitoring, environmental monitoring where multiple data is to be sent to sink node which is also known as Many-To-One communication model in which one sensor transmits different types of information to single base station node at single point of time, in this scenario, high accuracy is required. But as the density of sensor nodes increases, many-to-one type of communication will lead to high traffic load which can lead to energy depletion, significant increase in end-to-end delay and overall packet overhead and even packet loss. So, in order to maintain the overall quality of network, we propose an Energy Efficient Multipath Routing Algorithm based on Ant Colony Optimization to maintain energy level of nodes, avoid packet loss, routing overhead, less end-to-end delay and best throughput in the entire network.
3.3
Research Objectives
The following are the objectives of Research undertaken:
Analysis of Traditional Routing protocols as well as Ant Colony Optimization based Energy Efficient Routing Protocols to analyze the pros and cons with regard to routing issues, throughput, packet delivery and energy efficiency in sensor networks.
Comparison of ACO based routing protocols on basis of Energy Efficiency, Throughput, Packet Delivery Ratio, Routing Overhead and End to End Delay and shortlisting the best routing protocols on basis of pros and cons for further improvement.
Design and Development of Improvised Energy Efficient Routing Protocol based on Ant Colony Optimization to enhance overall performance of sensor network.
Performance comparison of proposed protocol with other existing ACO based routing protocols on basis of Packet Delivery Ratio, Throughput, End-to-End
79
Delay, Energy Efficiency and Routing overhead to determine the improvement in WSN network efficiency.
3.4
Research Methodology
In order to propose Improvised Energy Efficient Routing Protocol for Wireless Sensor Networks, the following methodology has been undertaken: Analysis
Comparison of Existing Protocols
Selection of Best Protocol
Design & Development of New Protocol
Validation and Testing with Existing Protocols
Checking the Results with existing protocols
Not OK
Redesign and Further Improvements in Protocol
If OK Draft of Final Thesis
Fig. 3.1. Research Methodology Undertaken
Detailed Literature review and analysis carried out for understanding the concept of sensor networks, applications, issues and challenges via Research Papers, Books, Forums and Online Research networks.
80
Detailed study of Traditional routing protocols as well as other novel protocols proposed by researchers is studied in depth to locate the pros and cons of routing and other issues in sensor networks.
Comparison of protocols is carried out on various parameters like Energy Efficiency, Throughout, Packet Delivery Ratio, Routing Overhead and End to End delay. Few routing protocols like AntChain, IACR, ACEAMRA, Basic ACO, EMCBR being shortlisted considering the pros and cons.
Design and Development of Improvised Energy Efficient ACO Based Routing Protocol is done via Swarm Intelligence- Ant Colony Optimization technique to enhance energy efficiency in sensor nodes and increasing the overall network lifetime of network.
The proposed protocol is simulated using NS-2 simulator on different simulation scenarios by taking different number of nodes and simulation time and hard core tested and compared with DSR, DSDV, Basic ACO, IACR, AntChain, EMCBR and ACEAMRA on basis of End-to-End delay, Routing Overhead, Energy Efficiency, Throughput and Packet Delivery ratio.
Analysis of data is being done and results stated that IEEMARP- Proposed protocol is almost 16% more efficient as compared to other existing routing protocols.
3.5 Research Contributions The Research contributions are as follows:
Analyzed various Energy Efficient Routing Protocols based on Ant Colony Optimization- Swarm Intelligence techniques in past literatures for Wireless Sensor networks along with their associated advantages and limitations. Determination of the problem, and finalization of research direction.
Analyzed and shortlisted the existing protocols to be taken as base for research and proposal of novel routing protocol i.e. Efficient Minimum-Cost Bandwidth
81
Constrained Routing (EMCBR), Basic Ant Routing Protocol (ACO), Ant Chain, Improved Ant Colony Routing (IACR) and ACO based Energy Aware Multipath Routing Algorithm (ACO-EAMRA).
Performed Simulations and Testing of protocols mentioned above using Mathematical formulas and working stated by respective authors using NS-2 simulator.
Designed and implemented a Novel Routing Protocol based on Ant Colony Optimization (Improvised Energy Efficient Multipath ACO Routing ProtocolIEEMARP for short) by overcoming all the drawbacks and limitations of existing protocols.
Evaluated the new proposed routing protocol i.e. IEEMARP in terms of Energy Efficiency, Packet Delivery Ratio, End-to-End delay, Throughput etc. and proved that IEEMARP routing protocol is more energy efficient and yields best network connectivity and performance and compared with other routing protocols like DSR, DSDV and Basic ACO.
3.6 Scope of Research The primary goal of this research work discussed in this thesis is to explore various Energy Efficient ACO based routing protocols for Wireless Sensor Networks. Various protocols are analysed on the basis of Energy Efficiency, Throughput, End-to-End delay, routing overhead and packet delivery ratio. The performance of network is analysed by deploying sensor nodes in dynamic changing environment on basic ant colony optimization routing protocol. It presents the simulation results by comparing the Basic ACO with DSDV, DSR routing protocols. The second goal of this thesis is to compare the shortlisted ACO based routing protocols like AntChain, ACOEAMRA, EMCBR and IACR and to observe the performance of network in varied parameters especially energy efficiency. By using Ant Colony Optimization based approach, a novel routing protocol i.e. IEEMARP is proposed which greatly minimizes the energy consumed in collecting and distributing the data among sensor nodes and back to sink node. IEEMARP routing protocol also outshines in terms of 82
packet delivery ratio, throughput, end-to-end delay and routing overhead and increases the overall network lifetime of wireless sensor network.
3.7 Research Gaps Identified As can be observed from the comprehensive literature review presented in Chapter 2, a large body of work exists in using Ant Colony Optimization based techniques for addressing various issues and challenges in Wireless Sensor Networks, including Energy Efficiency, Packet Delivery Ration, Throughput and End-to-End delay, there is still a research gap, and there is a need for further research efforts as many issues are still open and need to be solved. Some of the gaps are discussed as follows:
As Wireless sensor network nodes operate under resource constraint environments, network administrators face various design and implementation challenges especially with regard to Energy Efficiency, Packet delivery ratio and overall throughput. As, more than 85% of the energy of sensor nodes is consumed during communication operation (Sending and Receiving data in terms of packets), energy efficient techniques are needed to maintain throughout and overall energy among nodes to enhance network lifetime. Most of the previous approaches like AntChain, IACR, ACO-EAMRA, EMCBR routing protocols have various pros and cons and none of the routing protocol is efficient in maintaining energy efficiency, throughput and end-toend delay in overall network. Therefore, there is an emergent need of routing protocol based on ACO to provide proper energy load balancing, packet delivery ratio and overall throughput in the network.
As WSN sensors have limited resources, Swarm Intelligence based metaheuristic techniques are required for WSNs as compared to traditional routing algorithms. This will lead to less power computational resources and ACO based algorithm should be adaptive to dynamic changing topology, making nodes to learn changing environment. Existing ACO based algorithms have serious limitations to operate in less constrained environment as well as not adaptive and not fully efficient to operate in dynamic changing topology. So,
83
there is need of improvised ACO based routing protocol to take care of resource constraints and make the nodes adaptable to changing topologies.
Sensor network should be highly reliable network to ensure 100% delivery of packets from sender to sink node. Traditional routing protocols for WSN as well as existing ACO based routing protocols don’t have additional capability to provide acknowledgement of packet delivery. So, an emergent need of novel routing protocol is there to ensure accurate delivery of packets with proper acknowledgement of packet delivery.
84
Chapter – 4
IEEMARP: Improvised Energy Efficient Multipath Ant Colony Optimization based Routing Protocol for Sensor Networks This chapter covers the main contributions of this research undertaken. It describes the problem definition and includes the comprehensive description of Improvised Energy Efficient Multipath Ant Colony Optimization based Routing Protocol (IEEMARP) proposed for Wireless Sensor Networks for enhancing Energy Efficiency, Packet Delivery ratio, throughput, routing overhead and end-to-end delay in overall WSN network. The design and implementation of proposed routing protocol IEEMARP is also discussed.
4.1
Problem Definition and Background
As discussed in previous chapters, the state of art research is required in Wireless Sensor Networks. Wireless Sensor Networks are surrounded by various issues with regard to Route Optimality, Energy Efficiency, Security, Node Localization, Time Synchronization and many more. The most critical factor is routing protocol and it should be designed in such a manner which improvise the overall efficiency of sensor network. The traditional routing protocols for mobile Adhoc networks are classified as Proactive and Reactive, whereas, protocols designed for sensor network operations are categorized into Data Centric, Direct Diffusion, Hierarchical, Geographic routing [13]. Most of the traditional WSN routing protocols only consider network dynamics. In order to ensure consistency in networks, overhead increases with regard to increase in size of network traffic and network dynamics which in turn will write off network resources. Wireless Sensor Network routing protocols like Data Centric and Hierarchical are designed to attain optimality in data traffic with less energy utilization; but, the routing strategy especially in terms of path discovery is not highly taken care of in these protocols. Wireless Sensor Networks operations is quite tough, 85
a single node knows nothing when joins in the network, whereafter, limited energy is required to sense and update the topology. Under these situations, node failures, interferences and link disturbances will exist. To combat these situations and to operate in harsh environments, the WSN routing techniques require much more information that is exchanged by flooding throughout the network. In order to overcome all the situations, researchers across the world have made used use of various optimization techniques based on Genetic Algorithm, Fuzzy Logics, Evolutionary Computations, Soft Computing. To really optimize the routing and maintain energy efficiency in the network, Swarm Intelligence based techniques especially ACO is considered as baseline approach [35] [36]. In general, WSN routing protocol considers many parameters like packet transmission, sensing, node on-off status etc. Energy plays a significant role in heterogeneity environments in routing. Considering traditional Ant Algorithms, Ant Colony Optimization (ACO) protocol is used to select the routing path with high probability of pheromone. In case of malicious process, the ant releases the pheromone and the part of pheromone gets evaporated within certain time lapse. From this, the shortest routing path with high probability of pheromone is chosen. The existing ACO based techniques like AntNet, AdHocNet, Max-Min Ant System [45] [46] doesn‟t provide a perfect solution. Also, due to multipath routing, it contains source level path tables. Therefore, there is immediate need of new routing strategy, in which the energy verification and packet validation processes are considered. The level of energy is to be reduced to avoid unwanted packet transmission over control messages which in turn leads to the issue of Data Aggregation at the sink node. Also, the routing protocol needs to be improved so that it works on a heterogeneous based sensor network environment and concentrate on heterogeneity of sensing node. New systems are to integrate heterogeneity by considering various parameters of interesting nodes such as incentive, energy level, active state, latency and traffic rate. Further, predicting the interest of each node needs to be incorporated to reduce the unwanted level of energy consumption in node. If any node transmits the packet without interest, there is a 86
100% chance of getting the packet dropped or delayed in the process. Improvement in the latency rate of the packet delivery process by concentrating on interest of node needs to be considered. The analysis of parameters to get current pheromone level based on table driven multipath process is another option. During the process of routing, each node transmission processing is to be dynamically updated to the centralized router. In case of any packet transmission, router may use the node in the pre-selected routing path by using new path selection mechanism. Updating the node usage in the routing table and checking whether this routing path is affected or not is an important option need to be considered. In addition to above, message analysis should be considered. Based on this process, the unwanted node selection process for packet transmission is avoided. In addition to above, the routing protocol must consider all the parameters like Latency, Interest of nodes, Energy Consumption (Packet Generation, Alternate Path Selection, Packet Dropping, Trust Evaluation), Packet delivery ratio, throughput, delay and routing overhead avoidance.
4.2
Protocol Design Choices
Wireless Sensor Networks routing protocols should exhibit some necessary design features for successful operation of overall network. The features to be incorporated into a design will affect various aspects, namely energy efficiency. The main objective of the design of IEEMARP routing protocol is to incorporate those features which allow for an energy efficient routing protocol to adapt to the dynamic requirements of WSN environment. Listed below are some of the features that were selected for efficient operation of IEEMARP routing protocol.
4.2.1 Energy Efficiency The most important design criteria to be taken into account while designing any routing protocol for WSN environment is Energy Efficiency. Data transmission in wireless sensor node is the highest energy utilization factor and can take upto 80% of 87
node‟s power. This leads to the fact; the energy level should be optimized in such a manner that only those nodes which are required to transmit should WAKE up and transmit and rest all the nodes go to SLEEP mode.
4.2.2 Reliability In most of the real-world WSN bases applications, there are two main types of messages: Critical and Non-Critical. In order to test the reliability of sensor network, it is of utmost importance that all critical messages should be efficiently and timely delivered from source node to destination node. This in turn puts lots of pressure on the protocol to achieve this efficiently. With the use of Acknowledgement Messages (ACK) exchange between source node to destination node, the topology becomes reliable. The use of ACK messages for critical messages would be highly reliable solution. IEEMARP is highly reliable protocol making use of ACK messages and rechange of ACKR messages from request to response.
4.2.3 Dynamic Network and Scalability Wireless Sensor Networks are mostly operational in dynamic changing topologies and every time nodes gets failed or replaced or added in the network. For this, the routing protocol should be highly dynamic to understand the changing topology and scalable enough to adjust the newly added nodes in the network. IEEMARP is highly operational in dynamic environment and is robust and scalable enough to maintain overall scalability in the network.
4.2.4 Throughput and Routing Overhead Routing protocols should be highly efficient in maintaining overall network speed which maintains throughout in the network and reduces routing overhead by sophisticatedly balancing the traffic load. IEEMARP is designed to overcome this problem, as the protocol designed avoids packet loss, ensures time delivery of packets and reduces routing overhead which maintains network lifetime and agility.
88
4.3
Assumptions
In this research, we consider a sensor network consisting N sensor nodes which are randomly deployed in an environment. Sensor Nodes perform the task of collecting the data from the environment and transmit the data back to sink node. We suppose our Network Model and Network Protocol has been restricted by following assumptions:
Network Model:
All the nodes are homogeneous and are of equal size when deployed initially.
All the sensor nodes have the same amount of initial energy.
In the initial status, nodes don‟t have any information of any other neighbouring nodes like Location.
Every node in the sensor network is acting as Router, and has the capability to efficiently sense the surrounding nodes.
The speed of mobile nodes is random.
As the sensor network has issues with regard to energy, it forces to use the energy efficient algorithms in order to maximize the network lifetime.
WSN Nodes transceivers use single channel and wireless antenna, are omnidirectional and propagate isotropic signals in all directions.
WSN routing protocols makes a clear assumption to have in-advance knowledge of destination address.
Network Protocol :
The Protocol doesn‟t make use of broadcasting for any sort of route discovery and maintenance.
The protocol can have routing loops, but not persistent loops.
Energy Variations and Packet Delivery Ratio may occur in routing protocol as compared to other routing protocols i.e. ACEAMR, IACR, AntChain and EMCBR depending on applications and packet size.
Basic Ant Colony Routing Protocol is utilized for basic routing, energy maintenance and scenario working for sensor network.
89
4.4 IEEMARP Protocol- Operation Improvised Energy Efficient Multipath Ant Colony Optimization based Routing Protocol (IEEMARP) protocol operation comprise of three main phases. The phases are:
Neighbourhood Discovery via Link Knowledge
Forwarding of Packets / Fault Localization
Reliable End-to-End Communication from Source to Destination
4.4.1 Neighbourhood Discovery via Link Knowledge A sensor node sends a Hello message via one-hop broadcast to make its presence known to present nodes in its radio range. Nodes which are in the radio range are called neighbor nodes. A node may go to sleep mode to conserve energy level. When a node wakes up, it sends new hello packet is sent to all the neighbor nodes along with the duration of time it will be in active state in the message. If a node does not wish to be used for packet forwarding then it need not send a Hello message to its neighbors. Hello message allows the node to track of all its one-hop neighbors that are willing to forward packets and the duration the node is in active or awake state. The protocol doesn‟t enforce the node to monitor entire one-hop neighbors.
4.4.2 Forwarding of Packets / Fault Localization Let the subset of neighbors that the node keeps track be represented as N. Suppose a node s wants to transmit a packet, transmitted by another node, to a destination node d. For each neighbor node n ε N, the node s maintains a metric Rn,d, which represents the end to end reliability of forwarding a packet, going to destination node d through the neighbor node n. Initialization and updation of the Rn,d value are done by exponentially weighted moving average method which is discussed in the later section. If the destination node cannot be reached directly, then the node forwards the packet to an active and willing neighbor node. This neighbor node to which the data packet is 90
forwarded to will be henceforth referred to as the next-hop node. The next-hop node is chosen among the set of neighbors with highest R value. The data packets are not forwarded to the node from which it received the packet or to the source node which originated the packet. Every node in the network increments the hop-count value present in the header by one. The packet is discarded if the value of the hop-count exceeds a certain value, avoiding infinite loops in the network. IEEMARP protocol distributes the traffic load when multiple high quality routes exist between the nodes. IEEMARP is efficient enough to avoid congestions, packet delay and maintain link stability to avoid routing overhead and packet loss.
4.4.3 Reliable
End-to-End
Communication
from
Source
to
Destination The source node solicits an end to end acknowledgement packet known as ACK packet from destination node for packets that it transmits. The packets soliciting acknowledgement are called as ACK Request (ACKR) packets. The source node forwards these ACKR packets in a round robin fashion to all active and willing nodes. However the ACKR packets are subsequently treated in the same manner as data packets by all the other nodes to forward them to the next- hop node .i.e. the ACKR packets are treated similar to those of the data packets by being forwarded to the node with the highest R value. When ACKR paper is received, the destination node sends ACK packet to the source node. The ACK packet maintains end-to-end reliability value REL, initialized to 1 at destination node. On receiving the ACK packet originated by destination node d through neighbouring node n, the node updates the REL value as well as Rn,d . The methodology is demonstrated mathematically via following equations: ACK.REL = ACK.REL x macSucessRate
… (1)
Rn,d = (1-w) x Rn,d + w x ACK.REL
… (2)
91
MacSucessRate is regarded as current MAC-level success at the node and parameter is highly suitable. Furthermore, a node forwards ACK packet to the node through which it receives corresponding ACKR packet. So, ACK packet defines the end-to-end reliability for the route taken by ACKR packet originated from node s, forwarded by neighbouring node n and reaches the destination d. For every ACKR packet p a node generates or forwards, it needs to remember the packet‟s signature, sig(p), consisting of the source node, destination node, sequence number of the packet, the previous node and next hop node. The signature helps the ACK packet to be forwarded to the neighbor node from which it received the ACKR packet. The signature also helps the node to keep track of ACKR packets it received and eliminate any duplicate ACKR packets by discarding them. Every node after regular intervals of time, checks the quantity of signatures and removes signatures only if threshold value increases. Every ACKR packet maintains certain amount of end-to-end reliability of reaching the destination node d, through various traversed neighbouring nodes n, through the route taken by ACKR packet. Each ACK packet carries the current value of end-to-end reliability of reaching the destination node d through the neighbor node n along the route taken by the corresponding ACKR packet. The loss of a huge portion of the ACK packets by these transmissions is still acceptable as long as at least one ACK packet reaches the source node within a reasonable amount of time t (e.g. every 10 minutes). If no ACK packets are received in response to the ACKR packets forwarded to neighbor node n in time interval t. … (3)
Rn,d = (1-w) * Rn,d
The ACKR/ACK packets makes use of the highest reliability service that is provided by the MAC layer such as acknowledged transmission. 92
4.5 IEEMARP Routing Protocol- Properties Let N be the set of all the nodes in the network and D be the set of all the destination nodes in the network. Let η and Ө denote the cardinality of the sets N and D respectively. As the node need not keep track of all its one-hop neighbor nodes, η would be the upper limit of the number of neighbor nodes a node keeps track. The following things are required to be stored in node‟s memory:
A set N of one-hop neighbors that the node tracks.
Rn,d , n ε N and d ε D, will require O(ηӨ) memory.
{sig(p), n} , where sig(p) is the signature of an ACKR packet p forwarded to node n that has not yet reached the destination node and acknowledged.
Each node restricts the number of packet signatures that it can store to some threshold value. Therefore the total memory requirement of the node to store the packet signatures is O(η). The IEEMARP protocol can have routing loops, but these loops are not persistent. Suppose, we consider a scenario where a node A forwards a packet to another node B. Node B forwards this packet to node C, which in turn forwards this packet back to node A. By imposing the limit on the hop count value ensures that the packet will ultimately be discarded sometime once the hop count exceeds the limit. Also each node keeps track of the signature for the ACKR packet that it has forwarded to its neighbor nodes and has not yet been acknowledged yet. The node uses these signatures to discard any of the duplicate ACKR packets that might be travelling in the loop. Packets
forwarded
via
broadcast/multicast
cannot
utilize
the
MAC-level
acknowledgements and they are prone to loss due to collision at MAC-level and PHY-level noise. The IEEMARP protocol avoids broadcast/multicast for data packets forwarding or routing monitoring/discovery.
93
The protocol does not make use of broadcasting for route discovery and maintenance as it hinders the scalability of the network. One-hop broadcast is used for route discovery by sending hello packets to neighbouring nodes.
4.6 IEEMARP Protocol- Algorithm Input: Feature Matrix Output: Fitness Value Step 1: Initialize Population, )
√
))
)
)) //Where, x and y - input features of
node; Step 2: Initialize path, R = random value for size of feature matrix Step 3: Initialize Velocity, For i=1, V(i) = V + P (R(i), R(j)); End loop For k = 1 to number of cluster Omega, O(K+1) = max(omega) – (max(omega) – min(omega).max(R)); V (k+1) = O(k) * v(k) +
*(pb(k) – x(k)) +
* random * (Gb(k) – x(k));
//Where, pb – Path best; gb – Global Best; k- size of feature vector; b –number of updation; pd = trial intensity (k)* pb(path(k)); Step 4: Distance Calculation Step 5: Update PBest and GBest Step 6: Fitness value updating
94
Neighbor Discovery WSN Creation
1 0
2 1. Broadcast Hello Packet 2. Collecting Neighbor information
S
3
4
Route Discovery
S
Bant
Initialize
Fant
3
4
Create Ants Bant
5 Fant Construct Solution
D
Update global best Ph value S Termination Condition Yes
Initiate Packet Forward
Link Failure Bant Error Packet
Route Table
Data transmission Link State update
4
Fig. 4.1
IEEMARP Routing Protocol Operation
D
95
No
4.7
IEEMARP Routing Protocol- Algorithm
STEP 1: Suppose S being the source node want to transmit some data D being destination node with considering all QoS parameters like Energy Efficiency, faster transmission rate, avoidance of delay and quality bandwidth. The nodes visited by ants following the path from Node S to Node D are called Visiting nodes and list of visiting nodes is prepared. Visiting nodes list takes the form of multipath route table Rt. STEP 2: Node S is taken as initial node from where the transmission will start and will also initialize neighbor discovery process. STEP 3: Node S will initialize and transmit a Fant (Forward Ant- Route Request) to reach node D through via all path nodes at 1-hop distance from Node S. The Fant contains various parameters like address of source node, address of destination node, hop count and network speed. STEP 4: On pheromone evaporation, evaluation of 1-hop distance nodes will be performed. Each node „i‟ maintains a table called “ available pheromone on every link (
”, specifying the amount of
). This quantity is initialized to constant
C. )
Ph(i,j)=
… (1)
∑
s the amount of pheromone on the link. link visibility. α and β states the importance of pheromone to determine efficient route paths. M regarded as the set of
nodes
, not traversed by ant during packet
transmission. STEP 5: Calculation of pheromone evaporation of all 2-hop distance sensor nodes in the network.
96
STEP 6: The path preference probability value of each path from source S with the help of pheromone evaporation of every node. A node j from a set of adjacent nodes (j, k ... n) of i is selected as MPR node such that it covers all the 2-hop distance nodes and its path preference probability is better than others. STEP 7: If, path preference probability amount is higher as compared to predefined requirements, the path gets accepted and proceed with memory storage for effective utilization. STEP 8: On reaching the destination Fant gets converted to Bant. The Bant will take the same path like Fant in opposite way. STEP 9: The path having higher path preference probability will be considered as the optimal path for transmission of data packets from S source node to D destination node.
Summary ACO based routing protocols for sensor networks are highly dynamic, competent enough to understand topology and sensor node issues. IEEMARP routing protocol is based on Ant Colony Optimization technique to enhance network lifetime of sensor nodes via usage of forward and backward ant. Apart from energy efficiency, the routing protocol is highly reliable as protocol generates ACK packets from source to destination and vice versa to make sure of timely and accurate delivery of data packets.
97
Chapter – 5
Simulation and Performance Analysis of IEEMARP Routing Protocol This chapter describes the Simulation and Performance evaluation of IEEMARP Routing Protocol using performance metrics and simulation settings. The ultimate goal is to compare the performance of IEEMARP routing protocol with other ACO based routing protocols like ACEAMR, AntChain, EMCBR and IACR routing protocols.
5.1
Introduction to NS-2 Simulator
Network Simulator 2.35 (NS-2.35) [152, 153, 154, 155] is regarded as discrete event simulation tool and has proved its worth in research of dynamic communication networks. Thesimulator was developed in the year 1989 and since its inception various contributions are done that has brought various revolutions in the field of network research. The foundation of this simulator was based on REAL network simulator developed by University of California and Cornell University. NS-2.35 is totally based on Object Oriented (OO) programming so it is also known as Object Oriented Discrete Event Simulator [156, 157, 158, 159, 160]. It consists of two languages: C++ and Object Oriented Tool Command Language (OTcl). C++ is primarily used for implementing various protocols and extending simulation libraries whereas OTcl scripts does the task of configuring simulator, network topology setting, creating network scenarios and displaying simulation results. C++ and OTcl are binded together using TclCL. NS-2 comprise of 3,00,000 lines of code and is available free of cost and is used globally in academia. It can run on various operating systems like Linux, FreeBSD, MAC OS X, Solaris and even windows via use of third party software called Cygwin. The latest version is 2.36. 98
With regard to Wireless Sensor Network, NS-2 provides support for various protocols like 802.11, 802.16, 802.15.4, IR-UWB etc. But even though of lots of contributions from varied researchers around the world, NS-2 faces serious drawbacks in terms of WSN simulations like: Sensing model doesn‟t exist. The parameters which are used during simulation of nodes in WSN like energy model, packet formats, and MAC protocols are entirely different as we use in real world sensor network scenario. Another drawback surrounding NS-2 simulation results is that it also lacks application support which is required due to sensor network interaction between application level and protocol level. NS-2 supports dual output which can be either text-based or graphical based. For graphical based simulation NS-2 has inbuilt tool i.e. NAM (Network Animator) which shows live movement of packets in the nodes, node position and live simulation scenario and also contains XGraphs which shows the graphical analysis of the results drawn at the end of simulation.
Fig.5.1. Basic Structure of NS-2 Simulation
5.2
Performance Metrics
The following are the performance metrics which are used for evaluation of proposed protocol IEEMARP along with the comparison with other above mentioned ACO based routing protocols.
99
Energy Consumption: Energy consumption is regarded as battery power being utilized by each node for specific data transfer. This can be calculated by taking a difference of initial and final batter powers of each node. In a sensor network with „n‟ nodes, the following is regarded as the formula for calculating the total energy consumed. Ecosumed =∑
… (1)
Eix - Efx)
Eix= Energy of node x before transmission of data Efx= Energy of node x after transmission of data
Throughput: In sensor networks, throughput is the amount of digital data per time unit over a physical or logical link. It is measured in bits per second (bits/s or bps), occasionally in data packets per second or data packets per time slot. Throughput = Number of packers successfully transmitted / Total Time
… (2)
End-to-End Delay: End-To-End delay refers to the time taken for a packet to be transmitted across a network from source to destination. Data transmission seldom occurs only between two adjacent nodes, but via a path which may include many intermediate nodes. End-to-End delay is the sum of delays experienced at each hop from source to destination. End-To-End Delay = Time Spend on Hop 1 + Time Spend on Hop 2 + …+ … (3)
Time Spend on Hop n.
Packet Delivery Ratio: It is measure of the percentage of packets delivered successfully to the target nodes. Packet Delivery Ratio: (No. of Packets Sent – Packet Loss) * 100/no. of packets … (4)
sent.
Routing Overhead: It is the number of routing packets required for network communication. Routing overhead is calculated using awk script which processes the trace file and produces the final result. 100
5.3
Simulation and Performance Comparison of Basic Ant Colony Optimization (ACO) Routing Protocol with AODV, DSR and DSDV Routing Protocols for Wireless Sensor Networks
5.3.1 Flowchart of Simple Ant Net Based Routing Figure 5.2 demonstrates the flowchart of Simple Ant Net Based Routing.
Start
Nodes Deployment
Ant Route Discovery
Initiate New iteration
Find all new Routes
Solution Evaluation
If Dst is found?
Yes
No
Solution Found
End
Ph. placement
Ph. evaporation
Fig. 5.2. Flowchart demonstrating Ant Based Routing Algorithm
101
5.3.2
Simulation Parameters
Table 5.1. Outlines the Simulation Parameters- Basic ACO Routing Protocol Simulation Parameter
Value
NS2 Version
ns-allinone-2.35
Protocol
Ant Colony optimization(ACO) based Routing
Coverage Area
3000m X 1000m
Simulation Time
150,300,500
Antenna Type
Omni Antenna
Energy Model
EnergyModel(true)
Initial energy
10000mjoules
Number of Nodes
100,150,200…800
Queue Length
64
Data Rate
Variable
Interface Type
Wireless Physical Interface
Radio Range for Node
~250m
Mobility Speed
1…15m/s
Mobility Model
Random way point
5.3.3
Simulation Scenarios
Fig. 5.3. Deployment of 100 Nodes creating Wireless Sensor Network Scenario 102
In Figure. 5.3 above, 100 nodes are deployed in 3000m x 1000m forming Wireless Sensor networks and proper positioning of nodes is being done for simulating the routing of packets using ACO based routing protocol for determining the performance.
Fig. 5.4. Base Station location In above Figure. 5.4 “0” is determined as Base Station node located in the center of region for efficient data collection and aggregation from varied sensor nodes comprising WSN network.
Fig. 5.5. Route Discovery 103
In Figure. 5.5, the process of route discovery is initiated in which sensor node transmitting the data discover the neighbouring nodes from the base station node to transmit the data to destination.
Fig.5.6. Transmission of TCP Packet In Figure 5.6, after the discovery and estimation of effective routing path from source to destination using Ant Colony Optimization Algorithm, TCP packet is transmitted via relay node.
Fig. 5.7. Generation of Acknowledgement Packet 104
In Figure 5.7, after the transmission of TCP packet via relay nodes gets completed, Acknowledgement Packet is generated by Base station node and send to the source node to update the source node that packets are sucessfully received.
Fig. 5.8 Control Packet Overhead In Figure 5.8., TCP packet is dropped by sensor nodes due to control packet overhead.
Fig. 5.9. Handling of Route Failure by Ant Colony Optimization Algorithm 105
In Figure 5.9, as Wireless Sensor Nodes are highly mobile and operate autonomously in Dynamic Changing envoirnment. It has been observed as genuine nature that because of mobility, the TCP Packets are dropped and path is searched for next possible best path for data transmission. So, handling this type of situation can be effectively controlled by Ant Colony Optimization technique. 5.3.4
Simulation Results
In this simulation, we have compared Ant Colony Optimization Routing Protocol is compared with existing protocols of Wireless Sensor Networks. In order to prove the improvement in Routing provided by ACO based routing protocol, the results are compared with AODV, DSDV and DSR Routing Protocol on the basis of Probability of Packet Delivery Rate, Throughput, Routing Overhead, Energy Consumption and End-to-End Delay. a)
Packet Delivery Ratio
Packet Delivery Ratio based results are captured in Table 5.2 and depicted in Figure. 5.10. Table 5.2. Packet Delivery Ratio of ACO v/s DSDV, AODV and DSR Routing Protocols Packet Delivery Ratio (%age) Simulation Time DSDV
AODV
DSR
ACO
100
0.86
0.84
0.89
0.92
200
0.84
0.81
0.88
0.92
300
0.81
0.79
0.81
0.91
400
0.78
0.77
0.81
0.90
500
0.76
0.74
0.79
0.90
600
0.73
0.72
0.77
0.89
106
Packet Delivery Ratio (%age)
Packet Delivery Ratio 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 100
200 DSDV
300 AODV
400 DSR
500 ACO
Fig. 5.10. Packet Delivery Ratio comparing ACO with DSR, DSDV and AODV routing protocols. Considering Table 5.2 and Figure. 5.10 it is observed that ACO based solution has better packet delivery ratio almost 90% better as compared to AODV, DSDV and DSR considering simulation based data values. b)
Throughput
Throughput results are captured in Table 5.3 and depicted in Figure. 5.11. Table 5.3. Throughput analysis of ACO v/s AODV, DSDV and DSR routing protocols
Simulation Time
Throughput (Mbps) DSDV
AODV
DSR
ACO
100
6.23
8.21
8.02
9.58
200
7.05
8.14
9.12
10.26
300
9.23
9.68
10.36
11.23
400
9.1
10.25
11.23
12.32
500
9.86
10.14
10.36
11.69
600
9.87
9.76
10.02
11.45
107
Throughput (Mbps) 13
Throughput
12 11 10 9 8 7 6 5 100
200
300
DSDV
400
AODV
500 DSR
600 ACO
Fig. 5.11. Throughput comparing ACO with DSR, DSDV and AODV routing protocols Considering Table 5.3 and Figure. 5.11 it is observed that ACO based solution has better throughput almost 45% better as compared to AODV, DSDV and DSR routing protocols. c)
Routing Overhead
Routing overhead based data is captured in Table 5.4 and Figure. 5.12. Table 5.4. Routing Overhead based results and comparison of ACO with DSDV, AODV and DSR routing Protocols.
Simulation Time
Routing Overhead (%) DSDV
AODV
DSR
ACO
25
9.55
7.27
5.03
5.42
50
7.42
6.19
6.70
5.29
75
6.33
6.11
7.46
6.68
100
9.86
6.43
7.74
5.51
125
6.16
8.25
7.14
5.61
150
9.43
8.94
6.62
6.33
108
Routing Overhead (%age)
Routing Overhead 10 9.5 9 8.5 8 7.5 7 6.5 6 5.5 5 25
50
75
DSDV
100
AODV
125
DSR
150 ACO
Fig. 5.12. Routing Overhead reduced by ACO as compared to DSR, DSDV and AODV routing protocols It is analysed from Table 5.4 and Fig. 5.12, ACO based routing protocol makes less amount of routing overhead in overall network almost 35% less as compared to AODV, DSR and DSDV. d)
End-to-End Delay
End-to-End Delay based data is captured in Table 5.5 and Figure. 5.13. Table 5.5. End to End delay based results and comparison of ACO with DSDV, AODV and DSR routing Protocols.
Simulation Time
End to End Delay (Sec) DSDV
AODV
DSR
ACO
20
7.28
5.95
7.21
2.13
40
8.00
6.03
7.83
3.02
60
8.89
6.92
8.09
3.56
80
9.23
6.99
8.27
4.33
100
10.08
7.36
8.83
5.05
120
10.45
8.33
9.79
5.76
140
10.71
9.18
10.29
6.48
160
11.58
9.87
10.78
7.19
109
End To End Delay
End to End Delay (Sec.) 12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 20
40
60
DSDV
80
100
AODV
120 DSR
140
160 ACO
Fig. 5.13. End to End significantly reduced by ACO as compared to DSR, DSDV and AODV routing protocols From Table 5.5 and Figure. 5.13, it is analysed that ACO based Routing Protocol has significantly reduced End to End delay almost more than 40% as compared to AODV, DSR and DSDV routing protocols. e)
Energy Consumption
Energy consumption based data is captured in Table 5.6 and Figure. 5.14. Table. 5.6. Energy Consumption by Nodes in Path Selection and Data transmission between sender and receiver node. Number of Nodes
Energy Consumption (% mJoules) DSDV
AODV
DSR
ACO
25
98.21
98.2
99.42
99.52
50
97.23
97.25
98.32
98.21
75
93.26
94.25
94.68
95.23
100
88.42
89.21
90.25
91.45
125
85.05
86.33
87.59
89.06
150
81.17
82.79
84.30
86.20
175
77.30
79.26
81.01
83.33
200
73.43
75.72
77.72
80.47
110
Energy Conumption (% mJoules)
Energy Consumption 100 95 90 85 80 75 70 25
50 DSDV
75
100 AODV
125
150 DSR
175
200 ACO
Fig. 5.14. Energy Consumption by Node utilized via ACO protocol as compared to DSR, DSDV and AODV routing protocols It is observed that ACO based solution uses less energy as compared to traditional routing protocols like AODV, DSR and DSDV routing protocols. The reason behind ACO based routing protocol learns paths based on energy. Therefore, a path that leaves the nodes in the network with less amount of energy is selected whenever data transmission is required. This results in less energy consumption and maintains overall lifespan of network.
5.4
Simulation and Performance Comparison of Basic Ant Colony Optimization based Routing Protocols: ACEAMR, AntChain, EMCBR and IACR
5.4.1 Simulation Parameters The following Table 5.7 enlists the Simulation Parameters for comparison of ACEAMR, IACR, EMCBR and Ant Chain Routing Protocols.
111
Table 5.7. Simulation Parameters for Performance Comparison of ACEAMR, IACR, EMCBR and AntChain Routing Protocol. Simulation Parameters
Values
Simulator- Name and Version
Ns-allinone-2.35
Basic Routing Protocol
Modified AODV + Basic ACO Routing Protocol for Wireless Sensor Networks
Area
1000m x 1000m
Simulation Time
100 ms
Antenna Type
Omni-Directional Antenna
Energy Model
Energy Model (True)
Initial Energy- All Nodes in Network
100 J
No. of Nodes
100
Queue Length
50
Data Rate
Variable
Interface Type
Wireless Physical Interface
Radio Range of Nodes
250 m
Mobility Speed
Static
Mobility Model
Random Way Point
5.4.2
Simulation Results
a)
Packet Delivery Ratio
Table 5.8 and Figure. 5.15 highlights the Packet Delivery Ratio data captured via simulation of routing protocols.
112
Table 5.8. Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Packet Delivery Ratio
Simulation Time
Packet Delivery Ratio (%age) ACEAMR
AntChain
EMCBR
IACR
25
87.76
84.51
90.40
95.03
50
85.5
86.02
85.30
96.94
75
82.7
83.31
89.32
98.65
100
79.87
85.77
83.53
99.13
125
77.73
87
87.49
99.56
150
74.97
86.56
84.91
91.26
Packet Delivery Ratio 100 95
PDR(%)
90 85 80 75 70 65 60 25
50 ACEAMR
75 AntChain
100
EMCBR
125
IACR
150
Simulation Time (Seconds)
Fig. 5.15. Packet Delivery Ratio of ACEAMR, AntChain, EMCBR and IACR Analysis demonstrates that IACR is better as compared to other routing protocols almost 10-15% in performance. b)
Throughput
Table 5.9 and Figure. 5.16 highlights the Throughput data captured via simulation of routing protocols. 113
Table 5.9. Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Throughput
Simulation Time
Throughput (Mbps) ACEAMR
AntChain
EMCBR
IACR
25
198
182
175
190
50
223
213
187
182
75
217
212
205
207
100
189
187
185
213
125
193
188
185
184
150
178
188
210
204
Throughput(Mbps) 250
Throughput
200 150 100 50 0 25
50 ACEAMR
75 AntChain
100
125 EMCBR
150 IACR
Simulation Time (Seconds) Fig. 5.16. Comparison of Routing Protocols considering Throughput parameters Analysis state that EMCBR gives better throughput in Sensor Networks as compared to IACR, AntChain and ACEAMR routing protocol. c)
Routing Overhead
Table 5.10 and Figure. 5.17 highlights the Routing Overhead data captured via simulation of routing protocols. 114
Table 5.10. Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Routing Overhead Routing Overhead (%age) Simulation Time ACEAMR
AntChain
EMCBR
IACR
25
0.68
0.81
0.62
0.59
50
0.66
0.78
0.89
0.62
75
0.65
0.79
0.67
0.62
100
0.67
0.78
0.67
0.57
125
0.67
0.79
0.63
0.66
150
0.63
0.78
0.65
0.63
Routing Overhead (%age)
Routing Overhead 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 25
50 ACEAMR
75
100
AntChain
125 EMCBR
150 IACR
Simulaton Time (Seconds) Fig. 5.17. Comparison of Routing Protocols on basis of Routing Overhead Analysis that that ACEAMR and IACR performs better and less amount of routing overhead is observed during simulation of sensor networks as compared to AntChain and EMCBR routing protocol.
115
d)
Energy Consumption
Table 5.11 and Figure. 5.18 highlights the Energy Consumption by sensor nodes data captured via simulation of routing protocols. Table 5.11. Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of Energy Consumption Energy Consumption (% mJoules)
Energy Consumption (% mJoules)
Simulation Time ACEAMR
AntChain
EMCBR
IACR
25
98.21
98.2
98.81
99.42
50
97.23
97.25
97.785
98.32
75
85.05
86.33
86.96
87.59
100
81.17
82.79
83.54
84.30
125
77.30
79.26
80.13
81.01
150
73.43
75.72
76.72
77.72
Energy Consumption 100 95 90 85 80 75 70 65 60 25
50
ACEAMR
75
AntChain
100
125
EMCBR
150
IACR
Simulation Time (Seconds) Fig. 5.18. Energy Consumption comparison of Sensor Nodes on basis of different routing protocols in sensor networks 116
Analysis state that IACR is Best Protocol in Energy Conservation of nodes and is Highly Energy Efficient routing protocol as compared to ACEAMR, AntChain and EMCBR routing protocols. f)
End-to-End Delay
Table 5.12 and Figure. 5.19 states End-to-End delay data of sensor nodes captured via simulation of routing protocols. Table 5.12. Data Analysis of Routing Protocols (ACEAMR, AntChain, EMCBR and IACR) on basis of End-to-End Delay End to End Delay (Sec)
Simulation Time
ACEAMR
AntChain
EMCBR
IACR
20
67.12
68.19
48.28
50.10
40
66.89
45.16
57.53
66.47
60
38.82
60.82
46.39
44.24
80
44.25
47.25
39.43
42.59
100
67.55
64.98
41.77
66.52
120
56.37
60.16
43.03
47.98
140
64.67
52
46.78
67.75
End-To-End Delay
End To End Delay 80 70 60 50 40 30 20 10 0 20
40
ACEAMR
60
80
AntChain
100
EMCBR
120
140
IACR
Simulation Time (Seconds) Fig. 5.19. Comparison of Routing Protocols on basis of End-to-End Delay 117
Analysis state that AntChain is Best Protocol is End-to-End delay as compared to other routing protocols. 5.4.3
Overall Analysis and Best Protocol Suitability
Analysing all Simulation Scenarios and Metrics from the above Tables (Table 5.8 to Table 5.12) and figures (Figure 5.15 to Figure 5.19), the Table 5.13 states the best protocol chart in different metrics. Table 5.13. Overall Analysis of Best Protocol Suitability among ACEAMRA, EMCBR, IACR and AntChain Routing Protocol Network Parameter
Best Protocol
Packet Delivery
IACR
Throughput
EMCBR
Probability Routing Overhead
IACR, ACEAMRA
Energy Consumption
IACR
End to End Delay
AntChain
Overall, IACR is the Best Protocol among all other stated protocols for Wireless Sensor Networks.
5.5
Simulation and Performance Comparison of Proposed Routing Protocol
IEEMARP (Improvised Energy Efficient Multipath Ant Based Routing Protocol) with ACEAMR, Ant Chain, EMCBR and IACR routing protocols 5.5.1
Simulation Parameters
Table 5.14 enlists the Simulation Parameters taken for simulating novel proposed protocol i.e. IEEMARP and comparing it with ACEAMRA, AntChain, Basic ACO routing Protocol, IACR and EMCBR routing protocols.
118
Table 5.14. Simulation Parameters for Simulating IEEMARP Routing Protocol and Comparison with ACEAMRA, AntChain, IACR and EMCBR routing protocols Parameter Name
Values
Simulator Name and Version
ns-allinone-2.35
Base Protocol for Routing
Ant Colony Optimization (ACO) based Routing
Dimension of Topology
3000m x 1000m
Network Type
Wireless
Simulation Time
150s, 300s, 500s
Antenna Type
Omni Antenna
Simulation Model
Energy Model
Initial Energy of Nodes
10000mJ
Number of Nodes
100, 150, 200…800 (Max)
Queue Length
64
Data Rate
Variable
Interface Type
Wireless Physical Interface
Radio Range for Node
~250m
Mobility Speed
1….15m/s
MAC Type
IEEE 802.11
Mobility Model
Random way Point
5.5.2 Simulation Scenarios and IEEMARP Routing Protocol working In Figure. 5.20 NS-2 simulation is started and demonstrates the updating of neighbour nodes in routing table with pheromone values using basic ACO and modified AODV routing protocol. 119
Fig. 5.20. IEEMARP Simulation Start Scenario with Neighbouring Nodes Routing Table update using ACO routing protocol. In Figure. 5.21, the initial start of IEEMARP routing protocol is demonstrated. The IEEMARP protocol is performing the operation of Route Discovery to find neighbouring sensor nodes and updating the pheromone table.
Fig. 5.21. Route Discovery by IEEMARP Routing Protocol 120
In Figure 5.22, after determining the neighbouring sensor nodes and optimal routing paths, the sensor nodes will transmit the packets to Base Station (Sink Node- Node 0). The TCP based protocol is utilized to send the packets from sensor node to sink node for maintaining reliable end-to-end communication.
Fig. 5.22. Packet Transmission by Sensor Node to Sink Node via TCP Protocol for reliable end-to-end communication. In Figure. 5.23 after successful receipt of TCP packets from sensor node to base station, the base station will send the ACK Packet (Acknowledgement Packet) to the sensor node confirming the delivering making IEEMARP routing protocol highly reliable routing protocol.
121
Fig. 5.23 ACK Packet transmitted by Base Station (Node 0) to Sensor node confirming the acknowledgement of the received packet In Figure. 5.24, if any case any fault tolerance route discovery process, the packet is dropped because of routing overhead.
Fig. 5.24. Packet Dropped due to Routing Overhead 122
In Figure 5.25, the final link status between Source Node to Sink node is maintained at 10 Mbits/Sec at delay of 10 ms.
Fig. 5.25. Link Status between Source Node to Sink Node. 5.5.3
Performance Results of IEEMARP Routing Protocol with ACEAMR, AntChain, EMCBR and IACR routing protocols on performance metrics
The performance of IEEMARP routing protocol being proposed in this thesis is testing hard core in NS-2.35 simulator on different parameters and also compared with more than 5 routing protocols proposed by different researchers along with traditional routing protocols. This section will give detailed review of the results being analysed regarding performance of the routing protocol on different performance metrics to test the novelty and improvements in routing in sensor networks. a)
Packet Delivery Ratio
Table 5.15 and Figure. 5.26 states the comparative analysis of IEEMARP routing protocol with other protocols on basis of Packet Delivery Ratio.
123
Table 5.15. Performance comparison of Routing Protocols- ACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Packet Delivery Ratio Packet Delivery Ratio (%age) Time
ACEAMR
AntChain
EMCBR
IACR
IEEMARP
25
87.76
84.51
90.40
95.03
96.68
50
85.5
86.02
85.30
96.94
97.26
75
82.7
83.31
89.32
98.65
98.36
100
79.87
85.77
83.53
99.13
99.65
125
77.73
87
87.49
92.56
98.65
150
74.97
86.56
84.91
91.26
98.14
Fig. 5.26. Packet Delivery Ratio- Performance Metric Comparison It is analysed that IEEMARP outperforms in Packet Delivery Rate as compared to other routing protocols. IEEMARP performs almost 15% better in packet delivery among sensor nodes and maintains overall best service quality of packet delivery in sensor network.
124
b)
Throughput
Table 5.16 and Figure. 5.27 states the Throughput performance of IEEMARP routing protocol as compared to other routing protocols. Table 5.16. Performance comparison of Routing Protocols- ACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Throughput Throughput (Mbps) Time
ACEAMR
AntChain
EMCBR
IACR
IEEMARP
25
198
182
175
190
216
50
223
213
187
182
213
75
217
212
205
207
219
100
189
187
185
213
207
125
193
188
185
184
217
150
178
188
210
204
221
Fig. 5.27. IEEMARP throughput comparison with other routing protocols It is analysed that IEEMARP outshines in performance of throughput with almost 22% better in performance as compared to AntChain, ACEAMR, IACR and EMCBR protocol.
125
c)
Routing Overhead
Table 5.17 and Figure. 5.28 states the Routing overhead data results of IEEMARP routing protocol compared with other WSN routing protocols. Table 5.17. Performance comparison of Routing Protocols- ACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Routing Overhead Routing Overhead (%age) Time
ACEAMR
AntChain
EMCBR
IACR
IEEMARP
25
0.68
0.81
0.62
0.59
0.54
50
0.66
0.78
0.89
0.62
0.52
75
0.65
0.79
0.67
0.62
0.66
100
0.67
0.78
0.67
0.57
0.55
125
0.67
0.79
0.63
0.66
0.56
150
0.63
0.78
0.65
0.63
0.62
Fig. 5.28. Routing Overhead based Performance comparison of IEEMARP routing protocol with other routing protocols. 126
It is analysed that considering routing overhead, a serious issue reflecting the QoS of entire WSN network, IEEMARP routing protocol produces less routing overhead almost about 13% as compared to other routing protocols in sensor networks. d)
Energy Consumption
The most important aspect for any routing protocol is to maintain overall Energy Efficiency of nodes so that only those nodes can operate which are required to perform routing activity rest goes to sleep mode so that overall network lifetime can be enhanced. Table 5.18 and Figure 5.29 presents the Energy Consumption level of nodes after routing via IEEMARP protocol and compared with other routing protocols. Table 5.18. Performance comparison of Routing Protocols- ACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of Energy Consumption
Energy Consumption (% mJoules) Time
ACEAMR
AntChain
EMCBR
IACR
IEEMARP
25
98.21
98.20
98.81
99.42
99.52
50
97.23
97.25
97.785
98.32
98.21
75
85.05
86.33
86.95
87.59
89.06
100
81.17
82.79
83.54
84.30
86.20
125
77.30
79.26
80.13
81.01
83.33
150
73.43
75.72
76.72
77.72
80.47
127
Fig. 5.29. Energy Consumption comparison of Routing Protocols with IEEMARP routing protocol It is analysed that IEEMARP utilizes less Energy level of sensor nodes as compared to other routing protocols. The graphical results demonstrate that almost IEEMARP routing protocol is 8% more efficient as compared to other protocols. e)
End-to-End Delay
Table 5.19 and Figure. 5.30 demonstrates the review of End-to-End delay performance of IEEMARP routing protocol with other protocols.
128
Table 5.19. Performance comparison of Routing Protocols- ACEAMR, AntChain, EMCBR, IACR and IEEMARP on basis of End-to-End Delay End-To-End Delay (Sec) Time 20 40 60 80 100 120 140
ACEAMR 67.12 66.89 38.82 44.25 67.55 56.37 64.67
AntChain 51.99 42.06 67.35 65.30 55.11 60.95 49.13
EMCBR 48.28 57.53 46.39 39.43 41.77 43.03 46.78
IACR 50.10 66.47 44.24 42.59 66.52 47.98 67.75
IEEMARP 53.97 39.72 29.22 43.41 30.74 37.28 44.34
Fig. 5.30. End-to-End delay based performance comparison of IEEMARP routing protocol with WSN based other routing protocols Results and Graphs state that, IEEMARP maintains better End-to-End delay and maintains overall efficiency in network almost by 16% as compared to other routing protocols.
129
5.5.4
Performance
Comparison
of
IEEMARP
routing
protocol
with
Traditional WSN routing protocols: DSR, DSDV and Basic ACO Table 5.20 demonstrates the Data Values of Simulation based performance comparison on different evaluation performance metrics of IEEMARP routing protocol with other protocols of WSN. Table 5.20. Performance Comparison of IEEMARP routing protocol with DSDV, DSR and Basic ACO routing protocols on varied performance parameters Parameters
Basic ACO
DSDV
DSR
IEEMARP
Packet Delivery Ratio
94.58
95.96
88.12
97.73
Average End to End Delay
0.21
0.19
0.15
0.58
Average Number of Hops
1.67
1.092
1.04
8.21
Control Packet Overhead
9589
4809
5409
6090
Dropped Reply Messages
44
0
0
1
Throughput (Byes/s)
3012
2651
3094
3795
After analysing Table 5.20 on basis of varied performance parameters and comparing IEEMARP routing protocol performance with Basic ACO, DSDV and DSR routing protocol, IEEMARP has emerged as the Best Routing Protocol for maintaining overall efficiency in sensor network.
Summary IEEMARP routing protocol is tested and compared with Basic ACO, DSDV, DSR, IACR, AntChain, ACEAMR, EMCBR routing protocols on basis of Throughput, Packet delivery ratio, energy efficiency, routing overhead and end-to-end delay. The
130
protocol is tested on NS-2 simulator on different number of sensor nodes and simulation time and data analysis state that IEEMARP outshines in overall performance and maintains sensor nodes efficiency in better manner as compared to other routing protocols.
131
Conclusion and Future Scope Conclusion Energy Efficiency is regarded as highly critical issue to be addressed for Wireless Sensor Networks. In this thesis, an energy-efficient Ant-based Multipath Routing Protocol for WSNs (so-called IEEMARP) is proposed, and have compared its performance with ACEAMRA, AntChain, IACR, EMCBR, Basic ACO, AODV, DSDV and DSR routing protocols on basis of various performance metrics. For the sake of comparison, the same conditions have been introduced in route discovery process of well-known conventional AODV and BASIC ACO routing protocol, leading to Modified AODV and Improvised ACO routing protocol. Simulation based hardcore experiments are conducted to study the performance of proposed routing protocol on basis of Packet Delivery Ratio, Energy Efficiency, Throughout, Routing Overhead and End-To-End Delay. It has been observed that IEEMARP outshines in Energy Efficiency, reduces routing overhead, performs excellent throughput in overall network, improvises packet delivery ratio and reduces end-to-end delay to drastic level. IEEMARP is regarded as continuous learning protocol, where optimized paths are discovered and routes are stored in pheromone table for later use by sensor nodes for packet transmission. IEEMARP quickly adapts to Dynamic nature of network topology to search for optimal paths and make use of only those nodes with good energy level to the best possible extent. IEEMARP routing protocol consumes less energy of sensor nodes for data transfer request and also proven via data based simulation results to be the best proposed algorithm till date.
Future Scope As future work, the main efforts would be to strengthen the proposed IEEMARP routing protocol by making use of other Swarm Intelligence based techniques like Particle Swarm Optimization, Bee Colony Optimization, Bat Swarm or Elephant Swarm.
132
IEEMARP routing protocol would be enhanced more on Energy Efficiency, Security and End-to-End delay and like to compare our proposed protocol with other ACO based protocols like AntQHSen, FACOR, ANTLAG, FlockCC, Bat Swarm Algorithm etc. Efforts would be made to enhance IEEMARP routing protocol in terms of security by integrating various cryptographic techniques in terms of secured key exchange transfer during data transmission to make IEEMARP a reliable protocol. IEEMARP routing protocol would be enhanced to implement on real time sensor network kit for live environmental monitoring of data like Temperature, Humidity, Events and Accelerometer.
133
References 1.
Olariu, S., & Xu, Q. (2005, April). Information assurance in wireless sensor networks. In Parallel and Distributed Processing Symposium, 2005. Proceedings. 19th IEEE International (pp. 5-pp). IEEE.
2.
Akyildiz, I. F., Su, W., Sankarasubramaniam, Y., & Cayirci, E. (2002). Wireless sensor networks: a survey. Computer networks, 38(4), 393-422.
3.
Yick, J., Mukherjee, B., & Ghosal, D. (2008). Wireless sensor network survey. Computer networks, 52(12), 2292-2330.
4.
Chong, C. Y., & Kumar, S. P. (2003). Sensor networks: evolution, opportunities, and challenges. Proceedings of the IEEE, 91(8), 1247-1256.
5.
Ilyas, M., & Mahgoub, I. (Eds.). (2004). Handbook of sensor networks: compact wireless and wired sensing systems. CRC press.
6.
Sohraby, K., Minoli, D., & Znati, T. (2007). Wireless sensor networks: technology, protocols, and applications. John Wiley & Sons.
7.
Al-Karaki, J. N., & Kamal, A. E. (2004). Routing techniques in wireless sensor networks: a survey. IEEE wireless communications, 11(6), 6-28.
8.
Strazdins, G., Elsts, A., Nesenbergs, K., & Selavo, L. (2013). Wireless sensor network operating system design rules based on real-world deployment survey. Journal of Sensor and Actuator Networks, 2(3), 509-556.
9.
Cullar, D., Estrin, D., & Strvastava, M. (2004). Overview of sensor network. Computer, 37(8), 41-49.
10.
Tubaishat, M., & Madria, S. (2003). Sensor networks: an overview. IEEE potentials, 22(2), 20-23.
11.
Fahmy, H. M. A. (2016). Protocol Stack of WSNs. In Wireless Sensor Networks (pp. 55-68). Springer Singapore.
134
12.
Frey, H., Rührup, S., & Stojmenović, I. (2009). Routing in wireless sensor networks. In Guide to Wireless Sensor Networks (pp. 81-111). Springer London.
13.
Akkaya, K., & Younis, M. (2005). A survey on routing protocols for wireless sensor networks. Ad hoc networks, 3(3), 325-349.
14.
Vijayanand, S., & Suresh, R. M. (2007, December). An overlook on routing techniques in wireless sensor networks. In Information and Communication Technology in Electrical Sciences (ICTES 2007), 2007. ICTES. IET-UK International Conference on (pp. 940-945). IET.
15.
Krishnamachari, B., Estrin, D., & Wicker, S. (2002, June). Modelling datacentric routing in wireless sensor networks. In IEEE infocom (Vol. 2, pp. 3944).
16.
Intanagonwiwat, C., Govindan, R., & Estrin, D. (2000, August). Directed diffusion: A scalable and robust communication paradigm for sensor networks. In Proceedings of the 6th annual international conference on Mobile computing and networking (pp. 56-67). ACM.
17.
Di Caro, Gianni, and Marco Dorigo. "AntNet: Distributed stigmergetic control for communications networks." Journal of Artificial Intelligence Research 9 (1998): 317-365.
18.
Zhang, Y., Kuhn, L. D., & Fromherz, M. P. (2004, September). Improvements on ant routing for sensor networks. In International Workshop on Ant Colony Optimization and Swarm Intelligence (pp. 154-165). Springer Berlin Heidelberg.
19.
Patel, M., Chandrasekaran, R., & Venkatesan, S. (2004, June). Efficient Minimum-Cost Bandwidth-Constrained Routing in Wireless Sensor Networks. In International Conference on Wireless Networks (pp. 447-453).
20.
Ding, N., & Liu, P. X. (2004, August). Data gathering communication in wireless sensor networks using ant colony optimization. In Robotics and Biomimetics, 2004. ROBIO 2004. IEEE International Conference on (pp. 822827). IEEE. 135
21.
Peng, S., Yang, S. X., Gregori, S., & Tian, F. (2008, June). An adaptive QoS and energy-aware
routing algorithm
for
wireless sensor networks.
In Information and Automation, 2008. ICIA 2008. International Conference on (pp. 578-583). IEEE. 22.
Camilo, T., Carreto, C., Silva, J. S., & Boavida, F. (2006, September). An energy-efficient ant-based routing algorithm for wireless sensor networks. In International Workshop on Ant Colony Optimization and Swarm Intelligence (pp. 49-59). Springer Berlin Heidelberg.
23.
Xia, S., & Wu, S. (2009, November). Ant colony-based energy-aware multipath routing algorithm for wireless sensor networks. In Knowledge Acquisition and Modeling, 2009. KAM'09. Second International Symposium on (Vol. 3, pp. 198-201). IEEE.
24.
Garnier, S., Gautrais, J., & Theraulaz, G. (2007). The biological principles of swarm intelligence. Swarm Intelligence, 1(1), 3-31.
25.
Kennedy, J. F., Kennedy, J., Eberhart, R. C., & Shi, Y. (2001). Swarm intelligence. Morgan Kaufmann.
26.
Bonabeau, E., Dorigo, M., & Theraulaz, G. (1999). Swarm intelligence: from natural to artificial systems (No. 1). Oxford university press.
27.
Engelbrecht,
A.
P.
(2006). Fundamentals
of
computational
swarm
intelligence. John Wiley & Sons. 28.
Blum, C., & Li, X. (2008). Swarm intelligence in optimization. In Swarm Intelligence (pp. 43-85). Springer Berlin Heidelberg.
29.
Merkle, D., & Middendorf, M. (2014). Swarm intelligence. In Search methodologies (pp. 213-242). Springer US.
30.
Awad, M., & Khanna, R. (2015). Bioinspired computing: swarm intelligence. In Efficient Learning Machines (pp. 105-125). Apress.
31.
Dressler, F., & Akan, O. B. (2010). A survey on bio-inspired networking. Computer Networks, 54(6), 881-900.
136
32.
Beni, G., & Wang, J. (1993). Swarm intelligence in cellular robotic systems. In Robots and Biological Systems: Towards a New Bionics? (pp. 703-712). Springer Berlin Heidelberg.
33.
Engelbrecht, A. P. (2007). Computational intelligence: an introduction. John Wiley & Sons.
34.
Lim, C. P., & Dehuri, S. (Eds.). (2009). Innovations in swarm intelligence (Vol. 248). Springer.
35.
Yang, X. S., Cui, Z., Xiao, R., Gandomi, A. H., & Karamanoglu, M. (Eds.). (2013). Swarm intelligence and bio-inspired computation: theory and applications. Newnes.
36.
Fister Jr, I., Yang, X. S., Fister, I., Brest, J., & Fister, D. (2013). A brief review of nature-inspired algorithms for optimization. arXiv preprint arXiv:1307.4186.
37.
Bonabeau, E., Henaux, F., Guérin, S., Snyers, D., Kuntz, P., & Theraulaz, G. (1998). Routing in telecommunications networks with ant-like agents. Intelligent Agents for Telecommunication Applications, 60-71.
38.
Grassé, P. P. (1959). La reconstruction du nid et les coordinations interindividuelles chezBellicositermes natalensis etCubitermes sp. la théorie de la stigmergie: Essai d'interprétation du comportement des termites constructeurs. Insectes sociaux, 6(1), 41-80.
39.
Grassé, P. P. (1984). Termitologia, tome II. Fondation des sociétés. Construction.
40.
Millonas, M. M. (1992). Swarms, phase transitions, and collective intelligence (No. LA-UR-92-3980; CONF-9206329--1). Los Alamos National Lab., NM (United States).
41.
C. Blum and D. Merkle (eds.). Swarm Intelligence – Introduction and Applications. Natural Computing. Springer, Berlin, 2008.
137
42.
Panigrahi, B. K., Shi, Y., & Lim, M. H. (2011). Handbook of Swarm Intelligence. Series: Adaptation, Learning, and Optimization.
43.
Chu, S. C., Huang, H. C., Roddick, J. F., & Pan, J. S. (2011). Overview of algorithms for swarm intelligence. In Computational Collective Intelligence. Technologies and Applications (pp. 28-41). Springer Berlin Heidelberg.
44.
Dorigo, M. (1992). Optimization, learning and natural algorithms. Ph. D. Thesis, Politecnico di Milano, Italy.
45.
Dorigo, M., & Blum, C. (2005). Ant colony optimization theory: A survey.Theoretical computer science, 344(2), 243-278.
46.
Dorigo, M., Birattari, M., & Stutzle, T. (2006). Ant colony optimization. IEEE computational intelligence magazine, 1(4), 28-39.
47.
Stützle, T. (2009, April). Ant colony optimization. In International Conference on Evolutionary Multi-Criterion Optimization (pp. 2-2). Springer Berlin Heidelberg.
48.
Dorigo, M., Birattari, M., Blum, C., Clerc, M., Stützle, T., & Winfield, A. (Eds.). (2008). Ant Colony Optimization and Swarm Intelligence: 6th International Conference, ANTS 2008, Brussels, Belgium, September 22-24, 2008, Proceedings (Vol. 5217). Springer.
49.
Dorigo, M., & Stützle, T. (2003). The ant colony optimization metaheuristic: Algorithms, applications, and advances. In Handbook of metaheuristics (pp. 250-285). Springer US.
50.
Keller, L., & Gordon, E. (2009). The lives of ants. OUP Oxford.
51.
Karlson, P., & Lüscher, M. (1959). ‘Pheromones’: a new term for a class of biologically active substances. Nature, 183(4653), 55-56.
52.
Dorigo, M., & Gambardella, L. M. (1997). Ant colony system: a cooperative learning
approach
to
the
traveling
salesman
problem. Evolutionary
Computation, IEEE Transactions on, 1(1), 53-66. 53.
Dorigo, M., & Birattari, M. (2010). Ant colony optimization. In Encyclopedia of machine learning (pp. 36-39). Springer US. 138
54.
Dorigo, M. (2007). Ant colony optimization. Scholarpedia, 2(3), 1461.
55.
Dorigo, M., Maniezzo, V., & Colorni, A. (1996). Ant system: optimization by a colony of cooperating agents. Systems, Man, and Cybernetics, Part B: Cybernetics, IEEE Transactions on, 26(1), 29-41.
56.
M. Dorigo, T. Stützle, Ant Colony Optimization, MIT Press, Cambridge, MA, 2004.
57.
Dorigo, M., & Socha, K. (2006). An introduction to ant colony optimization.Handbook of approximation algorithms and metaheuristics, 26-1.
58.
Okafor, F. O., & Fagbohunmi, G. S. (2013). Energy Efficient Routing in Wireless Sensor Networks based on Ant Colony Optimization. West African Journal of Industrial and Academic Research, 8(1), 102-109.
59.
C.J. Watkins, P. Dayan (1992) Q-learning, Machine Learning, 8:279-292
60.
Stützle, T., & Hoos, H. H. (2000). MAX–MIN ant system. Future generation computer systems, 16(8), 889-914.
61.
Mirjalili, S. (2015). The ant lion optimizer. Advances in Engineering Software, 83, 80-98
62.
Zanjani, V. R., & Haghighat, A. T. (2009, November). Adaptive routing in ad hoc wireless networks using ant colony optimization. In Computer Technology and Development, 2009. ICCTD'09. International Conference on (Vol. 2, pp. 40-45). IEEE.
63.
Al-Zurba, H., Landolsi, T., Hassan, M., & Abdelaziz, F. (2011). On the suitability of using ant colony optimization for routing multimedia content over wireless sensor networks. International journal on applications of graph theory in wireless ad hoc networks and sensor networks, 3(2), 15-35.
64.
Mittal, P. K. Applications Of Ant Colony Optimization.
65.
Zengin, A., & Tuncel, S. (2010). A survey on swarm intelligence based routing protocols in wireless sensor networks. International Journal of Physical Sciences, 5(14), 2118-2126.
139
66.
Farooq, M., & Di Caro, G. A. (2008). Routing protocols for next-generation networks inspired by collective behaviors of insect societies: An overview. In Swarm Intelligence (pp. 101-160). Springer Berlin Heidelberg.
67.
Zungeru, A. M., Ang, L. M., & Seng, K. P. (2012). Classical and swarm intelligence based routing protocols for wireless sensor networks: A survey and comparison. Journal of Network and Computer Applications, 35(5), 15081536.
68.
Saleem, M., Di Caro, G. A., & Farooq, M. (2011). Swarm intelligence based routing protocol for wireless sensor networks: Survey and future directions. Information Sciences, 181(20), 4597-4624.
69.
Ali, Z., & Shahzad, W. (2011, July). Critical analysis of swarm intelligence based routing protocols in adhoc and sensor wireless networks. In Computer Networks and Information Technology (ICCNIT), 2011 International Conference on (pp. 287-292). IEEE.
70.
Di Caro, G., Ducatelle, F., & Gambardella, L. M. (2005, June). Swarm intelligence for routing in mobile ad hoc networks. In SIS (pp. 76-83).
71.
Zhang Y, Kuhn LD, Fromherz MPJ (2004). "Improvements on Ant Routing for SensorNetworks," Ant Colony, Optimization And Swarm Intelligence, Lecture Notes in Computer Science, 2004, 3172: 289-313.
72.
Mahale, R. A., & Chavan, S. D. (2013). AN OVERVIEW OF ANT COLONY BASED ROUTING PROTOCOLS IN WIRELESS SENSOR NETWORK.
73.
Wen, Y. F., Chen, Y. Q., & Pan, M. (2008). Adaptive ant-based routing in wireless sensor networks using Energy* Delay metrics. Journal of Zhejiang University SCIENCE A, 9(4), 531-538.
74.
GhasemAghaei, R., Rahman, A. M., Rahman, M. A., Gueaieb, W., & El Saddik, A. (2008, March). Ant colony-based many-to-one sensory data routing in wireless sensor networks. In 2008 IEEE/ACS International Conference on Computer Systems and Applications (pp. 1005-1010). IEEE.
140
75.
GhasemAghaei, R., Rahman, M. A., Gueaieb, W., & El Saddik, A. (2007, May). Ant colony-based reinforcement learning algorithm for routing in wireless sensor networks. In 2007 IEEE Instrumentation & Measurement Technology Conference IMTC 2007 (pp. 1-6). IEEE.
76.
Cai, W., Jin, X., Zhang, Y., Chen, K., & Wang, R. (2006, September). ACO based QoS routing algorithm for wireless sensor networks. In International Conference on Ubiquitous Intelligence and Computing (pp. 419-428). Springer Berlin Heidelberg.
77.
Wang, X., Li, Q., Xiong, N., & Pan, Y. (2008, October). Ant colony optimization-based location-aware routing for wireless sensor networks. In International Conference on Wireless Algorithms, Systems, and Applications (pp. 109-120). Springer Berlin Heidelberg.
78.
Ding, N., & Liu, P. X. (2005). A centralized approach to energy-efficient protocols for wireless sensor networks. In Mechatronics and Automation, 2005 IEEE International Conference (Vol. 3, pp. 1636-1641). IEEE.
79.
Misra, R., & Mandal, C. (2006). Ant-aggregation: ant colony algorithm for optimal data aggregation in wireless sensor networks. In Wireless and Optical Communications Networks, 2006 IFIP International Conference on (pp. 5-pp). IEEE.
80.
Zhu, X. (2007). Pheromone based energy aware directed diffusion algorithm for wireless sensor network. In Advanced Intelligent Computing Theories and Applications. With Aspects of Theoretical and Methodological Issues (pp. 283291). Springer Berlin Heidelberg.
81.
Sen, J., & Ukil, A. (2009, May). An adaptable and QoS-aware routing protocol for Wireless Sensor Networks. In Wireless Communication, Vehicular Technology, Information Theory and Aerospace & Electronic Systems Technology, 2009. Wireless VITAE 2009. 1st International Conference on (pp. 767-771). IEEE.
82.
Okdem, S., & Karaboga, D. (2009). Routing in wireless sensor networks using an ant colony optimization (ACO) router chip. Sensors, 9(2), 909-921. 141
83.
Guifeng, W., Yong, W., & Xiaoling, T. (2009, October). An ant colony clustering routing algorithm for wireless sensor networks. In Genetic and Evolutionary Computing, 2009. WGEC'09. 3rd International Conference on (pp. 670-673). IEEE.
84.
Ziyadi, M., Yasami, K., & Abolhassani, B. (2009, May). Adaptive clustering for energy efficient wireless sensor networks based on ant colony optimization. In Communication Networks and Services Research Conference, 2009. CNSR'09. Seventh Annual (pp. 330-334). IEEE.
85.
Wang, L., Zhang, R., & Geng, S. (2009, September). An energy-balanced antbased routing protocol for wireless sensor networks.
In
Wireless
Communications, Networking and Mobile Computing, 2009. WiCom'09. 5th International Conference on (pp. 1-4). IEEE. 86.
Jietai W, Jiadong XU, Mantian X. EAQR: An energy-efficient ACO based QoS routing algorithm in wireless sensor networks. Chinese Journal of Electronics 2009; 18: 113-6
87.
Guo, W., Zhang, W., & Lu, G. (2010, April). A comprehensive routing protocol in wireless sensor network based on ant colony algorithm. In Networks Security Wireless Communications and Trusted Computing (NSWCTC), 2010 Second International Conference on (Vol. 1, pp. 41-44). IEEE.
88.
Houshyarifar, V., & Amirani, M. C. (2014). Wireless Sensor Networks-A Review. International Journal of Computer Applications, 85(2).
89.
Baronti, P., Pillai, P., Chook, V. W., Chessa, S., Gotta, A., & Hu, Y. F. (2007). Wireless sensor networks: A survey on the state of the art and the 802.15. 4 and ZigBee standards. Computer communications, 30(7), 1655-1695.
90.
Arampatzis, T., Lygeros, J., & Manesis, S. (2005, June). A survey of applications of wireless sensors and wireless sensor networks. In Intelligent Control, 2005. Proceedings of the 2005 IEEE International Symposium on, Mediterrean Conference on Control and Automation (pp. 719-724). IEEE. 142
91.
Rashid, B., & Rehmani, M. H. (2016). Applications of wireless sensor networks for urban areas: a survey. Journal of Network and Computer Applications, 60, 192-219.
92.
Yuan, D., Kanhere, S. S., & Hollick, M. (2016). Instrumenting Wireless Sensor Networks—A survey on the metrics that matter. Pervasive and Mobile Computing.
93.
Rawat, P., Singh, K. D., Chaouchi, H., & Bonnin, J. M. (2014). Wireless sensor networks: a survey on recent developments and potential synergies. The Journal of supercomputing, 68(1), 1-48.
94.
Guo, W., & Zhang, W. (2014). A survey on intelligent routing protocols in wireless sensor networks. Journal of Network and Computer Applications, 38, 185-201.
95.
Akkaya, K., & Younis, M. (2005). A survey on routing protocols for wireless sensor networks. Ad hoc networks, 3(3), 325-349.
96.
Goyal, D., & Tripathy, M. R. (2012, January). Routing protocols in wireless sensor networks: A survey. In Advanced Computing & Communication Technologies (ACCT), 2012 Second International Conference on (pp. 474480). IEEE.
97.
Pantazis, N. A., Nikolidakis, S. A., & Vergados, D. D. (2013). Energyefficient routing protocols in wireless sensor networks: A survey. IEEE Communications surveys & tutorials, 15(2), 551-591.
98.
Rault, T., Bouabdallah, A., & Challal, Y. (2014). Energy efficiency in wireless sensor networks: A top-down survey. Computer Networks, 67, 104-122.
99.
Sharma, P., & Kaur, I. (2015). A Comparative Study on Energy Efficient Routing Protocols in Wireless Sensor Networks. International Journal of Computer Science Issues (IJCSI), 12(4), 98.
100.
Chu, S. C., Huang, H. C., Roddick, J., & Pan, J. S. (2011). Overview of algorithms for swarm intelligence. Computational Collective Intelligence. Technologies and Applications, 28-41.
143
101.
Fountas, C. (2010). Swarm Intelligence: The Ant Paradigm. In Multimedia Services in Intelligent Environments (pp. 137-157). Springer Berlin Heidelberg.
102.
Kordon, A. K. (2010). Swarm intelligence: The benefits of swarms. In Applying Computational Intelligence (pp. 145-174). Springer Berlin Heidelberg.
103.
Abraham, A., Guo, H., & Liu, H. (2006). Swarm intelligence: foundations, perspectives and applications. In Swarm Intelligent Systems (pp. 3-25). Springer Berlin Heidelberg.
104.
Parpinelli, R. S., & Lopes, H. S. (2011). New inspirations in swarm intelligence: a survey. International Journal of Bio-Inspired Computation, 3(1), 1-16.
105.
Blum, C. (2005). Ant colony optimization: Introduction and recent trends. Physics of Life reviews, 2(4), 353-373.
106.
Mohan, B. C., & Baskaran, R. (2012). A survey: Ant Colony Optimization based recent research and implementation on several engineering domain. Expert Systems with Applications, 39(4), 4618-4627.
107.
Blum, C. (2007, September). Ant colony optimization: introduction and hybridizations. In Hybrid Intelligent Systems, 2007. HIS 2007. 7th International Conference on (pp. 24-29). IEEE.
108.
Pei, Y., Wang, W., & Zhang, S. (2012, March). Basic ant colony optimization. In Computer Science and Electronics Engineering (ICCSEE), 2012 International Conference on (Vol. 1, pp. 665-667). IEEE.
109.
Shah, S., Bhaya, A., Kothari, R., & Chandra, S. (2013). Ants find the shortest path: a mathematical proof. Swarm Intelligence, 7(1), 43-62.
110.
Monteiro, M., Fontes, D. B., & Fontes, F. A. (2012). Ant Colony Optimization: a literature survey.
111.
Shtovba, S. D. (2005). Ant algorithms: theory and applications. Programming and computer software, 31(4), 167-178. 144
112.
Cordón García, O., Herrera Triguero, F., & Stützle, T. (2002). A review on the ant colony optimization metaheuristic: Basis, models and new trends. Mathware & soft computing. 2002 Vol. 9 Núm. 2 [-3].
113.
Dorigo, M. (2001, September). Ant algorithms solve difficult optimization problems. In European Conference on Artificial Life (pp. 11-22). Springer Berlin Heidelberg.
114.
Sim, K. M., & Sun, W. H. (2003). Ant colony optimization for routing and load-balancing: survey and new directions. IEEE Transactions on Systems, Man, and Cybernetics-Part A: Systems and Humans, 33(5), 560-572.
115.
Socha, K., & Dorigo, M. (2008). Ant colony optimization for continuous domains. European journal of operational research, 185(3), 1155-1173.
116.
Zhang, G., Pérez-Jiménez, M. J., & Gheorghe, M. (2017). Fundamentals of Evolutionary
Computation.
In Real-life
Applications
with
Membrane
Computing (pp. 11-32). Springer International Publishing. 117.
Yi, G., Jin, M., & Zhou, Z. (2010, June). Research on a novel ant colony optimization algorithm. In International Symposium on Neural Networks (pp. 339-346). Springer Berlin Heidelberg.
118.
Domínguez-Medina, C., & Cruz-Cortés, N. (2010). Routing algorithms for wireless sensor networks using ant colony optimization. Advances in soft computing, 337-348.
119.
Shirkande, S. D., & Vatti, R. A. (2013, April). Aco based routing algorithms for ad-hoc network (wsn, manets): A survey. In Communication Systems and Network Technologies (CSNT), 2013 International Conference on (pp. 230235). IEEE.
120.
Misra, S., Dhurandher, S. K., Obaidat, M. S., Gupta, P., Verma, K., & Narula, P. (2010). An ant swarm-inspired energy-aware routing protocol for wireless ad-hoc networks. Journal of systems and software, 83(11), 2188-2199. 145
121.
Kumar, N. A., & Thomas, A. (2012, July). Energy efficiency and network lifetime maximization in wireless sensor networks using improved ant colony optimization. In Computing Communication & Networking Technologies (ICCCNT), 2012 Third International Conference on (pp. 1-5). IEEE.
122.
Cheng, D., Xun, Y., Zhou, T., & Li, W. (2011). An energy aware ant colony algorithm for the routing of wireless sensor networks. Intelligent computing and information science, 395-401.
123.
Arya, R., & Sharma, S. C. Energy optimization of energy aware routing protocol and bandwidth assessment for wireless sensor network. International Journal of System Assurance Engineering and Management, 1-8.
124.
Zungeru, A. M., Seng, K. P., Ang, L. M., & Chong Chia, W. (2013). Energy efficiency performance improvements for ant-based routing algorithm in wireless sensor networks. Journal of Sensors, 2013.
125.
Wang, L., Sun, Q., & Ma, H. (2010, March). Energy consumption optimize based on ant colony algorithm for wireless sensor networks. In Informatics in Control, Automation and Robotics (CAR), 2010 2nd International Asia Conference on (Vol. 1, pp. 17-21). IEEE.
126.
Xia, S., Wu, S., & Ni, J. (2009, December). A new energy-efficient routing algorithm based on ant colony system for wireless sensor networks. In Internet Computing for Science and Engineering (ICICSE), 2009 Fourth International Conference on (pp. 176-180). IEEE.
127.
Almshreqi, A. M. S., Ali, B. M., Rasid, M. F. A., Ismail, A., & Varahram, P. (2012, February). An improved routing mechanism using bio-inspired for energy balancing in wireless sensor networks. In Information Networking (ICOIN), 2012 International Conference on (pp. 150-153). IEEE.
128.
Jiang, X., & Hong, B. (2010, June). ACO based energy-balance routing algorithm for WSNs. In International Conference in Swarm Intelligence (pp. 298-305). Springer Berlin Heidelberg.
146
129.
Orojloo, H., Moghadam, R. A., & Haghighat, A. T. (2012). Energy and path aware ant colony optimization based routing algorithm for wireless sensor networks. Global trends in computing and communication systems, 182-191.
130.
Lin, Y., Zhang, J., Chung, H. S. H., Ip, W. H., Li, Y., & Shi, Y. H. (2012). An ant colony optimization approach for maximizing the lifetime of heterogeneous wireless sensor networks. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 42(3), 408-420.
131.
Hui, X., Zhigang, Z., & Xueguang, Z. (2009, July). A novel routing protocol in wireless sensor networks based on ant colony optimization. In Environmental Science and Information Application Technology, 2009. ESIAT 2009. International Conference on (Vol. 2, pp. 646-649). IEEE.
132.
Arabshahi, P., Gray, A., Kassabalidis, I., Das, A., Narayanan, S., Sharkawi, M. E., & Marks, R. J. (2001). Adaptive routing in wireless communication networks using swarm intelligence.
133.
Laxmi, V., Jain, L., & Gaur, M. S. (2006, December). Ant colony optimization based routing on NS-2. In
International Conference on Wireless
Communication and Sensor Networks (WCSN), India. 134.
Ali, Z., & Shahzad, W. (2013). Analysis of routing protocols in ad hoc and sensor wireless networks based on swarm intelligence. International Journal of Networks and Communications, 3(1), 1-11.
135.
Ahmed, M. B., Boudhir, A. A., & Bouhorma, M. (2012). New routing algorithm based on ACO approach for lifetime optimization in wireless sensor networks. International Journal of Networks and Systems, 1(2).
136.
II, D. D. R. P. (2010). A Routing Protocol Based on Ant Colony Algorithm for Wireless Sensor Networks. Chinese Journal of Electronics, 19(4).
137.
Norouzi, A., Babamir, F. S., & Zaim, A. H. (2011). A novel energy efficient routing protocol in wireless sensor networks. Wireless Sensor Network, 3(10), 341.
147
138.
Ahvar, S., & Mahdavi, M. (2011). EEQR: An energy efficient query-based routing protocol for wireless sensor networks. Journal of Advances in Computer research, 2(3), 25-38.
139.
Gajjar, S., Sarkar, M., & Dasgupta, K. (2014). Self Organized, Flexible, Latency and Energy Efficient Protocol for Wireless Sensor Networks. International Journal of Wireless Information Networks, 21(4), 290-305.
140.
Lin, T. L., Chen, Y. S., & Chang, H. Y. (2014, August). Performance Evaluations of an Ant Colony Optimization Routing Algorithm for Wireless Sensor Networks. In Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP), 2014 Tenth International Conference on (pp. 690693). IEEE.
141.
Guo, H. (2012, June). Investigation on ant-colony based routing algorithm for wireless sensor networks. In World Automation Congress (WAC), 2012 (pp. 227-229). IEEE.
142.
Tong, M., Chen, Y., Chen, F., Wu, X., & Shou, G. (2015). An energy-efficient multipath routing algorithm based on ant colony optimization for wireless sensor networks. International Journal of Distributed Sensor Networks.
143.
Saleem, K., Fisal, N., Hafizah, S., Kamilah, S., & Rashid, R. A. (2009). Ant based self-organized routing protocol for wireless sensor networks. International Journal of Communication Networks and Information Security, 1(2), 42.
144.
Wang, Z., & Zhang, D. (2005, November). An improved ACO algorithm for multicast routing. In IFIP International Conference on Network and Parallel Computing (pp. 238-244). Springer Berlin Heidelberg.
145.
Singh, A., & Behal, S. (2013). Ant colony optimization for improving network lifetime in wireless sensor networks. International Journal of Engineering Sciences, 8, 1-12.
148
146.
Li, Z., & Shi, Q. (2013, November). An QoS Algorithm Based on ACO for Wireless
Sensor
Network.
In
High
Performance
Computing
and
Communications & 2013 IEEE International Conference on Embedded and Ubiquitous Computing (HPCC_EUC), 2013 IEEE 10th International Conference on (pp. 1671-1674). IEEE. 147.
Shen, Y., Pei, Q., Xu, Q., Feng, H., & Ma, J. (2009, December). A routing algorithm based on ant-colony in wireless sensor networks. In Computational Intelligence and Security, 2009. CIS'09. International Conference on (Vol. 2, pp. 441-445). IEEE.
148.
Kate, V. B., & Das, S. (2014). Energy Efficient Ant Colony Optimization based Routing Protocol for Wireless Sensor Networks.
149.
Vergados, D. J., Pantazis, N. A., & Vergados, D. D. (2007, August). Enhanced route selection for energy efficiency in wireless sensor networks. In Proceedings of the 3rd international conference on Mobile multimedia communications (p. 63). ICST (Institute for Computer Sciences, SocialInformatics and Telecommunications Engineering).
150.
Sahni, J. P. S., & Park, J. (2005). Maximum Lifetime Routing in Wireless Sensor Networks. Computer & Information Science & Engineering, University of Florida June, 2.
151.
Sundani, H., Li, H., Devabhaktuni, V., Alam, M., & Bhattacharya, P. (2011). Wireless sensor network simulators a survey and comparisons. International Journal of Computer Networks, 2(5), 249-265.
152.
Kellner, A., Behrends, K., & Hogrefe, D. (2010). Simulation environments for wireless sensor networks. no. June, 2.
153.
Egea-Lopez, E., Vales-Alonso, J., Martinez-Sala, A. S., Pavon-Marino, P., & García-Haro, J. (2005, July). Simulation tools for wireless sensor networks. In Summer Simulation Multiconference, SPECTS (pp. 2-9).
154.
Fall, K., & Varadhan, K. (2005). The ns Manual (formerly ns Notes and Documentation). The VINT project, 47.
149
155.
The Network Simulator-ns2-Available from: http://www.isi.edu/nsnam/ns
156.
Issariyakul, Teerawat, and Ekram Hossain. Introduction to network simulator NS2. Springer, 2011.
157.
http://www.nsnam.org/docs/release/3.14/tutorial/singlehtml/index.html (Accessed on May 3, 2017)
158.
Discrete Event Simulation: URL http://en.wikipedia.org/wiki/Discrete_event_ simulation#Network_simulators”. Accessed on 17 August, 2014.
159.
Kellner, A., Behrends, K., & Hogrefe, D. (2010). Simulation environments for wireless sensor networks. no. June, 2.
160.
Nayyar, A., & Singh, R. (2015). A comprehensive review of simulation tools for wireless sensor networks (WSNs). Journal of Wireless Networking and Communications, 5(1), 19-47.
150
