Simulation Study of FELGossiping Protocol Performance for Active Security System Kais Mekki, Ahmed Zouinkhi, Mohamed Naceur Abdelkrim Research Unit MACS (Modeling, Analysis and Control of Systems) National Engineering School of Gabes Gabes, Tunisia
[email protected],
[email protected],
[email protected] Abstract—One of the most important issues in Wireless Sensor Networks (WSN) is collecting and processing data perceived from the environment and sending that data to be processed and evaluated. So routing data towards the destination node is a fundamental task in WSN. However, the sensor nodes have a limited transmission range, and their processing and storage capabilities as well as their energy resources are also limited. Therefore, different routing protocols have been developed for WSN. In this paper, we evaluate the performance of FELGossiping routing protocol for active security system of dangerous chemical products warehouse. We compare its performance with AODV, SPEED and EQSR protocols in terms of packet delivery ratio, end to end delay and energy consumption. Simulation results using Castalia/OMNeT++ show that FELGossiping has good packet delivery ratio and energy consumption performances, and medium average end to end delay. Keywords—Wireless Sensor FELGossiping, Castalia.
Network,
Routing
protocols,
I. INTRODUCTION In recent years, the rapid technological advances in low power and highly integrated digital electronics, small scale energy supplies, tiny microprocessors, and low power radio technologies have created low cost and multifunctional wireless sensor devices, which can observe and react to changes of their surrounding environments. These sensor devices are equipped with a small battery, a tiny microprocessor, a radio transceiver and a set of sensor modules that used to acquire information that reflect the changes in the surrounding environment of the node. A typical Wireless Sensor Networks (WSN) consists of a number of sensor devices that collaborate with each other to accomplish a common task (e.g. environment monitoring and object tracking) and report the collected data through wireless interface to a center node (the sink). The areas of applications of WSN vary from civil, healthcare and environmental to military. Examples of applications include target tracking in battlefields, habitat monitoring, civil structure monitoring, forest fire detection and factory maintenance.
978-1-4799-0313-9/13/$31.00 ©2013 IEEE
However, with the specific consideration of the properties of sensor networks such limited power, stringent bandwidth, dynamic topology (due to node failures, adding/removing nodes or even physical mobility), high network density and large scale deployments have posed many challenges in the design and management of WSN. These challenges have demanded energy awareness and robust protocol designs at all layers of the networking protocol stack. For these reasons, different protocols are designed for routing purpose in WSN [1]. Efficient utilization of sensor energy resources and maximizing the network lifetime were and still are the main design considerations for the most proposed routing protocols and algorithms for sensor networks. However, depending on the type of application, the generated sensory data have different attributes, where it may contain delay sensitive and delay tolerant data [2]. For example, the data generated by a sensor network that monitors the temperature in a normal weather monitoring station are not required to be received by the processing center or the sink node within certain time limits. On the other hand, for a sensor network that is used for fire detection in a forest, any sensed data that carries an indication of a fire should be reported to the processing center within certain time limits. In our paper, we aim to present a routing algorithm for an active security system of dangerous chemical products warehouse. The purpose of the system [3] [4] is to monitor chemical warehouse because such storage management products may cause great danger if safeguards are not respected. In this application, every chemical container is equipped with a wireless sensor node which controls the internal state of the product and the external changes of its environment (e. g. temperature, brightness and humidity) using different sensors modules. The node also control the distance between the containers by periodic exchange of greeting messages between neighboring nodes (RSSI method) to prevent the closeness of the products that are not compatible. If there are critical or unusual events (e. g. high temperature, incompatible
135
products are very close), the node is able to make decisions and send alert packets to control centerr via the sink. In addition, the nodes periodically exchange configuration and information messages with the control cennter to monitor the temporal evolution of the chemical productss. This security system uses a centralized approach a and pointto-point connection for the communication between nodes and sink. The nodes communicate directly withh the control center by Broadcast Model [5] (Application Layerr + Data link Layer + Physical Layer). For a large scale waarehouse and high distance, the node will be unable to com mmunicate with the control center due to the low transmission power of wireless sensor nodes. So, the node must greeatly increase its transmission power to reach the sink whicch consuming huge amount of energy and greatly limits the lifeetime of nodes and hence the network lifetime. To overcome this limitation, we proposed in this paper a multi-hopss connection for nodes/control_center communication. Thee messages passes from one node to another until it reaches the control center, this mechanism requires a routing protocol. Furthermore, the application generates security alerts. An alert is transferred by a high priority packet across the wireless network to achieve the sink, hence this real-time application requires strict constraints on both delay annd loss packets rate in order to report the alert data to the conntrol center within certain time limits without any loss. Thereffore, enabling realtime communication in sensor networks reequires energy and QoS awareness on network layer in orderr to have efficient utilization of the network resources and effective e access to sensors readings. Thus, our attention shoould focus on the energy efficient and QoS based routing prottocols. QoS routing is an important topic inn sensor networks research, and it has been the focus of the reesearch community of WSN. Many QoS based routing prootocols specifically designed for WSN have been proposed, forr example, SPEED [6], AODV [7], RPAR [8], THVR [9], FELGossiping [10], RRR [11], EQSR [12], and MARP [13]. For our chemical security system, we had chosen Fair Efficieent Location-based Gossiping (FELGossiping) which is one off the recent routing algorithms. In this paper, we compare FELGossipping with AODV, SPEED and EQSR which are respectively considered as flat QoS based protocol, location QoS based protocol and multipath QoS based protocol. These protocols have been implementted and simulated using Castalia/OMNeT++ Tools languagee and compared in terms of packet delivery ratio, end to endd delay and energy consumption to evaluate the performance off FELGossiping. The rest of the paper is organized as follows. Section 2 provides a description of the active securityy system. Section 3 presents the different categories of routing protocols in WSN. Gossiping protocol. Section 4 gives a description of FELG Section 5 presents our implementation and a our simulation results. Section 6 concludes the paper.
CURITY SYSTEM II. ACTIVE SEC
Accidents in the chemical inndustry are becoming frequent due to the absence of adequate security measures especially in the field of the storage and haandling of dangerous chemical products. This topic has attracted the interest of several research projects. The application of our workk has been done to meet the needs of this field [3], its goal g is to monitor dangerous chemical products warehouse, because the management of such products may cause a greaat danger if the safety rules are not respected. This application was able to turn the dangerous products in a communicatingg entity (active product) by integrating a sensor platform in every chemical container with a dangerous substance, thereforre, upgrading it with interaction capacities in the middle of itts action environment. If two active products are in the samee proximity, they communicate between them through messaages sent by radio frequency waves. So, an active producct can communicate with its environment, make decisions and a send information to a sink node which interfaces the wireeless sensor network to control center Ethernet as shown in figuure 1 [4]. WIFI
Sink
Ethernet Control center
WSN
Figure 1. Active security system of dangerous chemical products warehouse
The sensors glued to chemicaal units must periodically send information about the status off the products (e.g. temperature and humidity). If there is an abrupt change from an environmental or internal sttate of dangerous chemical products, the application at the sensors s must report this alert to the control center by high prioriity packets. The security system uses a centralized c approach and pointto-point connection for the com mmunication between nodes and control center as shown in figure 2. For a large scale warehouse and high distance, thhe node will be unable to reach and communicate with the coontrol center due to the low transmission power of wireless sensor nodes. To overcome this limitation, we aims to usee a multi-hops connection for nodes/control_center communiccation through routing protocol as shown in figure 3.
136
Control center
nodes toward the sink [17]. Multipath based: In this category, the network derives benefit from the fact that there are multiple paths between a source node and the sink. Using different paths enhance the fault tolerance and ensure that the nodes energy is consumed uniformly [18].
Product 2
Product 1
Product 3
Product 4
Figure 2. Point-to-point connection for nodes/control_center communication Control center
Product 1
Product 2
QoS based: QoS based protocols have to find a trade-off between energy consumption and the quality of service. A high energy consumption path may be adopted if it improves the QoS. So when interested in energy conservation, these types of protocols are usually not very useful [19]. Heterogeneity based: In heterogeneity sensor network architecture, there are two types of sensors namely linepowered sensors which have no energy constraint, and the battery-powered sensors having limited lifetime, and hence we should use their available energy efficiently through routing protocol by minimizing their potential of data communication and computation [15].
Product 3
Product 4
Figure 3. Multi-hops connection for nodes/control_center communication
III. ROUTING PROTOCOLS IN WSN Routing in WSN differs from conventional routing in traditional networks in various ways. There is no infrastructure, wireless links are unreliable, sensor nodes may fail, and routing protocols have to meet strict energy saving requirements [14] [15]. Many routing algorithms were developed for WSN. All major proposed routing protocols may be divided into seven categories: Flat based: In this category, all nodes have the same role and there is no hierarchy. Flat routing protocols distribute information as needed to any reachable node within the sensor transmission range. No effort is made to organize the network or its traffic, only to discover the best route hop by hop to a destination by any path [14]. Hierarchical based: This class of routing protocols sets out to conserve energy by arranging the nodes into clusters. Nodes in a cluster transmit data to a head node within close proximity which aggregates the collected information and forward data to the sink [16]. Location based: Most of the routing protocols for WSN require location information for sensor nodes. In most cases, location information is needed to calculate the distance between two particular nodes so that energy consumption can be estimated (transmission power). Since there is no addressing scheme for sensor networks like IP-addresses, location information can be utilized in routing data in an energy efficient way [14].
IV. WORKING OF FELGOSSIPING FELGossiping routing protocol [10] improve the problems of Gossiping and its extensions (SGDF, LGossiping and ELGossiping). FELGossiping consists of three phases: Network Initialization Phase, Information Gathering Phase and Routing Phase. A. Network Initialization Phase The network initialization phase starts after the sensor nodes are randomly distributed in the controlled area. In the beginning, the sink node broadcasts a HELLO message to its neighbors. The HELLO message contains the hop count which is used to setup the node distance to the sink (gradient). After broadcasting the HELLO massage, all 1-hop neighbors will receive this message. Each node saves the hop count in its memory and increases it by 1. The new hop count is then replaced with the old one. After each node has received the HELLO message it will continue to broadcast this message to farther nodes. When a node receives a HELLO message, it will check whether it already has a gradient. If it has a gradient, it will compare it to the hop count of its own message and will replace it with the message’s hop count if the latter is smaller, and will add 1 to the hop count prior to broadcasting it. However, if its gradient is smaller than or equal to the hop count of the message, it will drop the message. As a result, the gradient will keep the best route. Finally, the process will continue until all the nodes receive the HELLO message, at that time the network initialization phase will be completed and each node through the gradient knows its distance to the sink. Figure 4 shows the flowchart of this initialization phase.
Mobility based: Mobility brings new challenges to routing protocols in WSN. Mobile nodes require energy efficient protocols to guarantee data delivery originated from source
137
information gathering phase and the routing phase are shown in the flowchart in figure 7.
Start
Sink node prepares HELLO message
Broadcast HELLO message
All nodes receive the message?
YES
Finish
S
S
(b) Step 2
(a) Step 1
Figure 5. Source node (S) selects the neighbor that had the most residual energy
NO Receive HELLO message
YES
Is there a gradient?
Control Center
NO
Sink S
Check the hop count of the message
Smaller? NO
YES
Save the hop count of the message, increment it by 1 and put back the new hop count in the message.
Drop the message
Figure 6. Routing phase Start
Send Request message
Figure 4. Flowchart of the network initialization phase
B. Information Gathering Phase After detecting the event, the source node broadcast a request message to acquire the information from the neighboring nodes in its transmission range. The nodes that received the request message send their information to the source node, the response message contains the hop count and the residual energy of the neighboring node. C. Routing Phase After the information gathering phase finishes, the routing phase will start. The source chooses within its transmission range two neighboring nodes that have the minimum hop count towards the sink as shown in figure 5(a). Then, the source compare between these two nodes according to the residual energy, the nodes that had the most residual energy will be the next hop as shown in figure 5(b). If two nodes have the same residual energy, the nodes that have a lower hop count to the sink will be taken. After that, the source node sends the packet to the selected node. Upon receiving the packet, the node as next hop repeats the information gathering phase and routing phase to transmit the message to another node as shown in figure 6. The process continues until the packet reaches the sink or the TTL is finished. The
Receive Response messages
Choose two nodes with minimum hop count to Sink
Compare the residual energy of these two nodes and choose the highest one to be the next hop
Send the packet
Sink node?
YES
Finish
NO TTL = TTL - 1
TTL = 0 ?
YES
Drop the packet
NO Figure 7. Flowchart of information gathering phase and routing phase
138
V. PERFORMANCE EVALUATION In this section, we describe firstly the simulation setup. Then, we discuss the simulation results. A. Simulation Setup To implement the routing protocols we have used Castalia/OMNeT++ simulator. Currently many wireless sensor network simulators are available as NS2 and SENSE but Castalia provides realistic wireless channel and radio models, and realistic node behavior especially relating to access of the radio [20] [21]. The simulation parameters that we have chosen are summarized in table 1, and have been selected so as to be compatible with other simulation studies of WSN [6] [12] [22].
Figure 9. Packet delivery ratio (in percentage % form)
TABLE I. SIMULATION SETUP
Simulation platform Castalia version Network area Number of sensor nodes Sensor distribution Location of Sink Radio range Number of source nodes Data rate of source nodes MAC protocol
Linux Ubuntu 10.04 LTS 3.2 800m*800m 500 Uniform random Center of area 20 m 100 3 packet/s TMAC
Initial energy of nodes
18720 J
Radio
Consistent with Telos components (CC2420)
Figure 10. Average end to end delay (in ms)
Figure 8 shows the nodes positions that are randomly distributed within an 800m x 800m square (m=meters). The communication range is 20 meters, and the sink is located at the center of the square.
Figure 11. Energy consumption performance
Figure 8. Nodes deployment
B. Simulation Results To evaluate the performance of FELGossiping we compared its performance with AODV, SPEED and EQSR. Figures 9, 10 and 11 present the performance of FELGossiping, AODV, SPEED and EQSR in terms of packet delivery ratio, average end to end delay and energy consumption, respectively.
Figures 9 and 10 show that the performance of FELGossiping in terms of packet delivery ratio is very good approximate 91% and better than AODV and SPEED, so we can say this protocol is applicable to carry sensitive information in WSN but it fails for the scenario where transmission time should be very low as it has high end to end delay. FELGossiping suffers from a lost time on the information gathering phase despite it always chooses the paths that have the minimum number of hops to the sink. In this phase and for each received packet, the protocol broadcasts a request and waits for response messages from all the neighboring nodes to determine the next hop which causes more queuing and treatment delay for received packets.
139
Despite this, the information gathering phase is very useful for mobile chemical containers management of our security system as it discover neighboring nodes in each hop. Furthermore, this phase could detect the failed nodes for static containers through neighbors table which is updated in every neighboring discovery. So, the node from the old table which is absent in the new one, is considered as failed and therefore the source node must initiate alert packet which contain the ID and the location of the absent node and send it to the control center. However, average end to end delay of FELGossiping is better than On Demand routing protocol (AODV). The EQSR protocol performs better than FELGossiping, AODV and SPEED in terms of packet delivery ratio and average end to end delay as it use multipath algorithm and dividing the packet into a number of segments and using multiple paths to route segments simultaneously to the sink. Figure 11 shows that the FELGossiping protocol consumes the lowest energy and therefore ensures a long network lifetime than others QoS based protocols AODV, EQSR and SPEED because it select the node that had the most residual energy as next hop among the closest nodes to the sink. So, FELGossiping is an energy efficient algorithm and guarantee high packet delivery ratio. Furthermore, this protocol is very useful for containers mobility management and nodes failure detection through the information gathering phase as it discover neighboring nodes in each hop. However, we have to enhance the end to end delay performance of this algorithm to route alert packet within a very short time. VI. CONCLUSION In this paper we have evaluated FELGossiping routing algorithm for chemical products security system and we compared its performance with AODV, SPEED and EQSR. These protocols have been evaluated on Castalia/OMNeT++ simulator by using three metrics, packet delivery ratio, average end to end delay and energy consumption. The packet delivery ratio and energy consumption metrics have shown that FELGossiping has good reliability and is energy efficient algorithm. But, the protocol has medium quality of service on average end to end delay metric. However, FELGossiping could be very useful for mobile chemical containers management and nodes failure detection of our active security system. As a future project, we will try to improve the end to end delay of FELGossiping for routing of packets to the control center within certain time limits. REFERENCES [1] [2] [3]
K. Romer and E. Mattern, "The design space of wireless sensor networks," IEEE Wireless Communications, Vol. 11, No. 6, December 2004, pp. 54-61. V.C. Gungor and G.P. Hancke, “Industrial wireless sensor networks: challenges, design principles and technical approaches”, IEEE Transactions on Industrial Electronics, Vol. 56, 2009, pp. 4258–4265. A. Zouinkhi, E. Bajic, E. Rondeau and M. N. Abdelkrim, “Simulation and modeling of active products cooperation for active security system
[4]
[5]
[6]
[7]
[8]
[9]
[10] [11] [12] [13]
[14] [15] [16]
[17]
[18] [19]
[20] [21]
[22]
management”, International Journal Transactions on Systems, Signals and Devices, Vol. 5, No. 3, 2011, pp. 1-23. A. Zouinkhi, E. Bajic, E. Rondeau, M. Ben Gayed and M. N. Abdelkrim, “Ambient Intelligence Awareness context application in industrial storage”, Journal of Wireless Sensor Network, Vol. 3, 2011, pp. 134-145. D. Dobre, E. Bajic, A. Zouinkhi, “Active product modeling based on smart object concept: application to chemical security management”, European Journal of Automated Systems, Vol. 43, No. 4, 2009, pp. 561579. T. He, J. Stankovic, C. Lu and T. Abdelzaher, “SPEED: A stateless protocol for real-time communication in sensor networks", 23rd International Conference on Distributed Computing Systems, 19-22 May 2003, pp. 46–55. Z. Che-aron, W. Al-Khateeb and F. Anwar “An Enhancement of AODV routing protocol for more robust wireless sensor network”. International Journal of Computer and Network Security, Vol. 2, No. 7, 2010, pp. 1826. O. Chipara, Z. He, G. Xing, Q. Chen, X. Wang, C. Lu, J. Stankovic and T. Abdelzaher, “Real-time power-aware routing in sensor network”, 14th IEEE International Workshop on Quality of Service, 19-21 June 2006, pp. 83–92. Y. Li, C.S. Chen, Y. Song, Z. Wang and Y. Sun, “Enhancing Real-Time Delivery in Wireless Sensor Networks with Two-Hop Information”, IEEE Transactions on Industrial Informatics, Vol. 5, No. 2, May 2009, pp. 113-122. A. Norouzi, F. S. Babamir and A. H. Zaim, “A Novel Energy Efficient Routing Protocol in Wireless Sensor Networks”, Journal of Wireless Sensor Network, Vol. 3, 2011, pp. 350-359. E. Gelenbe and E. Ngai, “Adaptive QoS Routing for Significant Events in Wireless Sensor Networks”, 5th IEEE International Conference on Mobile AdHoc and Sensor Systems, September 2008, pp. 410–415. J. B. Othman and B. Yahya, “Energy efficient and QoS based routing protocol for wireless sensor networks”, Journal of Parallel and Distributed Computing, Vol. 70, No. 8, August 2010, pp. 849-857. J. Sen and A. Ukil, “An Adaptable and QoS-Aware Routing Protocol for Wireless Sensor Networks”, 1st International Conference on Wireless Communication, Vehicular Technology, Information Theory and Aerospace & Electronic Systems Technology, May 2009, pp. 767–771. K. Akkaya and M. Younis,“A Survey on Routing Protocols for Wireless Sensor Networks”, Ad-hoc Networks, Vol. 3, No. 3, May 2005, pp. 325349. L.J. García Villalba, A.L. Sandoval Orozco, A.T. Cabrera and C.J. Barenco Abbas, “Routing Protocol in Wireless Sensor Networks”, Sensors 2009, Vol. 9, No. 11, pp. 8399-8421. S.K. Singh, M.P. Singh and D.K. Singh, “A survey of Energy-Efficient Hierarchical Cluster-based Routing in Wireless Sensor Networks”, International Journal of Advanced Networking and Application, Vol. 2, No. 2, September 2010, pp. 570–580. X. Zhang, J. He and Q. Wei, “Energy-Efficient Routing for Mobility Scenarios in Wireless Sensor Networks”, Proceedings of the 3rd International Symposium on Electronic Commerce and Security Workshops, Guangzhou P. R. China, 29-31 July 2010, pp. 80-83. M. Radi, B. Dezfouli, K. A. Bakar and M. Lee, “Multipath Routing in Wireless Sensor Networks: Survey and Research Challenges”, Sensors 2012, Vol. 12, No. 1, pp. 650-685. J. Chen , M. Diaz, L. Llopis, B. Rubio and J. M. Troya, “A survey on quality of service support in wireless sensor and actor networks: Requirements and challenges in the context of critical infrastructure protection”, Journal of Network and Computer Applications, Vol. 34, No. 4, 2011, pp. 1225-1239. A. Boulis, “Castalia, a simulator for wireless sensor networks and body area networks, version 3.2”, User’s manual, NICTA, March 2011. H. Sundani, H. Li, V. Devabhaktuni, M. Alam and P. Bhattacharya, “Wireless Sensor Network Simulators, A Survey and Comparisons”, International Journal of Computer Networks, Vol. 2, No. 5, 2011, pp. 249-265. L. Hey and E. Gelenbe, “Adaptive Packet Prioritization for Wireless Sensor Networks”, Proceedings of the 5th Euro-NGI Conference on Next Generation Internet Networks, Aveiro Portugal, 1-3 July 2009, pp. 63-69.
140
