2011 International Conference on Communication Systems and Network Technologies
Enhanced Reliable Distributed Energy Efficient Protocol for WSN Shamneesh Sharma#1, Robin Prakash Mathur#2, Dinesh Kumar#3 Department of Computer Science & Engineering Lovely Professional University- Punjab (India)
[email protected]#1,
[email protected]#2,
[email protected]
Abstract: - This paper presents a concept of solving the back-up node failure problem of the RDEEP (Reliable Distributed Energy Efficient Protocol). This technique will also help to make the wireless sensor network more energy efficient. The energy efficiency is one of the most important factors in case of wireless sensor networks. In this era of green computing, ERDEEP (Enhanced Reliable Distributed Energy Efficient Protocol) provides more energy efficient network and makes it more reliable as compared to the RDEEP. This scheme is based upon the principle that the cluster head will choose at least three backup nodes on the basis of best Nodial distances from each other. In case of failure of the back-up node new backup node will perform all the functions of the cluster head. Keywords: - Reliable Distributed Energy Efficient Protocol, Primary Backup Node, Secondary Back-up Node, Cluster Head Selection, Back-up Node Selection, Stable Election Protocol, Distributed Energy Efficiency Protocol and LEACH.
I. INTRODUCTION Wireless sensor network [1] is the emerging area of research in present days. Researchers are trying to overcome the drawbacks for the existing field and improving it day by day so that it can be used in the service of mankind. The Wireless Sensor Network is a network of the MEMS [2] (MicroElectro-Mechanical-System) which has three characteristics Self computation, communication and sensing capabilities. Sensor nodes mainly depend upon battery for power and most of it is very difficult to recharge. So data processing and routing scheme should be energy saving, before they are put to use with sensor node. . In the beginning of the technology the sensor node deployment of sensors was limited as we can deploy on few sensor node due to size and cost factors but due to the evolution of the ULSI (Ultra Large Scale Integration) [3] technology now we are able to deploy thousands of node in the sane field. Due to this technology no one think less than 500 nodes to deploy in the field. In the beginning of the technology the homogeneous devices were used in the sensor fields; then there was incessant rain of ideas for energy efficiency after the invention of clustering technique. So many protocols have been used in the field of energy efficiency. In the same series LEACH [4] (Low Energy Adaptive Clustering Hierarchy) protocol comes in to 978-0-7695-4437-3/11 $26.00 © 2011 IEEE DOI 10.1109/CSNT.2011.41
picture and the huge range of maxims has been drawn in the same series. The trend of the heterogeneous wireless sensor networks started from the invention of SEP [5] (Stable election Protocol). These networks were more energy efficient. SEP was the one of the most elegant improvements in the LEACH. It was heterogeneous scheme in which all nodes did not have same electrical energy and processing capabilities. II. RELATED WORK There are so many protocols invented by various researchers in the field of the energy efficiency of wireless sensor networks. All the protocols were focusing upon the energy factor but reliability was the issue that was untouched. In the series of reliability RDEEP [6] (Reliable Distributed Energy Efficiency Protocol) comes in to picture. This protocol solves the problem of the reliability by choosing a Back-up Node for every cluster head. The Back-up Node will remain in the sleep mode and if cluster head gets down; a signal is send to back-up node along with data and responsibilities by cluster head to back-up node and it (Back-up node) starts its working by taking all the responsibility of the cluster head. Now this back-up node will aggregate the data and forward to the base station. But this protocol has the problem of Single Point of Failure. As there is probability that the Backup Node gets down also; in that case the reliability factor fails. Now in this case the RDEEP fails to attempt the responsibility of the cluster head to forward the data of the different nodes to the base station. After some time when the growth in this field gets its pace many protocols have been invented which decrease the data redundancy and were more energy efficient [7, 8 & 9]. We define the whole network consisting of three types of nodes as active, sleeping, passive and dead nodes. As we know that there can be by taking these four types of nodes into consideration we just focus on the following points: 1) The protocol is made reliable by keeping the first nearest nodes to the cluster head in sleep and two more nearest nodes as passive nodes. The sleeping node gets automatically active when cluster head sends an activation signal to it when it (cluster head node) reaches at the specified threshold value of energy. Rest of two nodes is the secondary back-up nodes for the cluster head and become active when Back-up node sends them activation signal. 2) As there are three nodes in each cluster in which one is in sleep mode and two are in the passive mode so the network
162
will be more energy efficient and prolong life span of the network make sure the process of Green computing [10].
2. Primary Back-up Node Selection
III. ARCHITETURE OF ERDEEP RDEEP is the enhanced version of DEEP [11] (Distributed Energy Efficiency Protocol) which overcame the drawback of the Reliability up to some extent ,so in the same fashion the ERDEEP is the enhanced version of the RDEEP which removes the back-up node failure problem form this protocol. The whole concept of ERDEEP moves around the RDEEP. This protocol consists of four parts as described below: 1. Cluster Head Selection As it has been defined in the RDEEP to select the cluster head and our protocol also based on the same so same formula is used for the purpose of the cluster head selection: .
1
.
.
.
Where k= No. of cluster
The node which will be at the minimum Nodial Distance from the Cluster Head Node will be elected as the Primary Back-up Node and this node becomes sleeping node. The following formula will be used to calculate the Nodial Distance of the sensor nodes: -
Where (x1, x2) are the co-ordinates of cluster head and (y1, y2) are the co-ordinates of the simple node with reference to the cluster head. 3. Secondary Back-up Node Selection The secondary Back-up node will be selected as per the same rule but the next two nodes having more than equal to Nodial Distance to the Back-up node. Both of the nodes are selected as Secondary Back up nodes and that is why they become passive nodes. 4. Activation of Primary and Secondary Nodes The Back-up Node will take the responsibility of the Cluster head if it gets down. When the cluster head reaches at certain defined threshold value of energy it sends activation signal to the Primary Back-up node and makes it as active node now it will perform all the functioning of the cluster head. The Secondary Back- up nodes will become active if the Back-up Node becomes dead node
L= L bit Message .
IV. WORKING OF THE WHOLE SYSTEM
[5] In this protocol there is assumption that m be the fraction of the total number of nodes n, which are equipped with α times more energy than the others. The powerful nodes are referred as advanced nodes, and the rest (1 − m) × n as normal nodes. The energy requirement for the non cluster head is given by the following formula: .
.
.
The whole system will follow some steps from the starting to the end of the network. The steps are described below: 1) In the beginning of the process we will deploy the sensor nodes at the desire place. The node contains some advanced nodes along with the simple nodes. 2) If the all node are of homogeneous energy then deploy more sensor nodes with extra capabilities to make the network as heterogeneous one. 3) Now communication starts with in the nodes.
Here
4) The cluster head formation and choosing the cluster head is chosen as per energy distribution among the network. So we can say that the total energy for one time communication in the cluster can be described as below:-
5) After the election of the cluster head, it broadcast a declaration signal to all of its members.
The total energy for one time communication can be described as below: -
6) After receiving the declaration signal the members send their acknowledgement to cluster head along with the acceptance signal and their Nodial Coordinate points. These coordinate points are calculated with reference to the coordinates of the cluster head.
. 2
.
2
7) Cluster head calculate the Nodial Distance of its entire member group. 163
8) Now Cluster head select three most near nodes as primary and secondary Back-up nodes and set primary node in sleep mode and other two in passive mode. 9) After the completion of first round of data aggregation, Cluster Head check its Threshold energy limit (T (e)). If it is less than threshold value it will be lesser than the specified one, it will send the signal to BKP and set it as active node. 10) After this the Back-up node become the cluster head and performs all the functioning of the cluster head. 11) Now after sending aggregated data to the base station, it checks its threshold energy value. If it is less than equal to (T (e)) it sends the activation signal to the secondary nodes and set one as active node and one as sleep node. 12) After processing next round started prom the cluster formation and election of new cluster heads. V. ALGORITHM FOR ERDEEP A. Initial phase
2) CH {Cal (ND)} of each member as per following formula: -
3) CH {Sort (ND1, ND2, ND3, ND4, ND5, ND6, ND7, ND8, ND9, ND10)}. 4) CH {Set node with Min (ND) as PBKP} in this case ND2=Min ND. Set PBKP= SLEEP. 5) CH {Set next two nodes with Min (ND) as SBKP} Set SBKP= PASSIVE. D. Initialization of Primary Backup node 1) If CH (e) Send value of their (x, y) coordinates to CH. C. Selection of primary and secondary back up nodes 1) If {n (x, y)}>=N where N= Number of members in the cluster.
2) Set SBKP1 = ACTIVE.
4) PBKP: Send data to SBKP1. 5) If SBKP1 (e)
