Australian Journal of Basic and Applied Sciences, 5(9): 1376-1380, 2011 ISSN 1991-8178 (EBHCR) Energy Balancing and Hierarchical Clustering Based Routing algorithm for Wireless Sensor Networks 1 Roghaiyeh Gachpaz Hamed, 2 Jaber Karimpour. 1 Computer Engineering Department Islamic Azad University, Qazvin,Iran 2 Faculty of Electrical & Computer Engineering University of Tabriz, Tabriz, Iran Abstract: Whereas the energy constraint is one of the most important restrictions in wireless sensor networks so the energy balancing is essential for prolonging the network lifetime. In the recent years many protocols have been proposed for clustering on wireless sensor network to balance the energy consumption. This paper proposes a new clustering protocol namely EBHCR which uses hierarchical dynamic routing algorithm. This algorithm selects a best sensor node in terms of energy, number of neighbors and distance as a cluster head. Simulation results show that the EBHCR prolongs the network lifetime about 45% compared to the LEACH. Key words: Wireless sensor network; clustering; hierarchical routing; energy balancing. INTRODUCTION With the incredible advances in micro electromechanical systems (MEMS) and wireless communication, use of wireless sensor networks (WSNs) has considerably increased. Wireless sensor networks are containing of the thousands or more sensors that are widely distributed in the environment. Distribution of the sensors in the environment can be done manually or randomly. These networks have many applications such as environmental monitoring, healthcare, military operations, target tracking and etc. Due to the power constraint, energy balancing and maximizing network lifetime have been important challenge in wireless sensor networks. For this reason, use of data aggregation which limits redundant transmission between sensors is essential. One of the techniques that use the data aggregation to reduce energy consumption in WSNs is called clustering based routing algorithm (Lin &Gerla 1997; Hou & Tsai 2001; Ryu et al., 2001; Jiang & Manivannan 2004). In (Heinzelman et al., 2002; Heinzelman et al., 2000; Younis & Fahmy 2004; Soro & Heinzelman 2005; Younis & Fahmy 2004; Rasouli Heikalabad et al., 2010) there are protocols that use the clustering technique in the network. LEACH is one of the most famous clustering based routing protocols in WSN (Heinzelman et al., 2000). Cluster head selection among sensor nodes is done randomly and also data transmitting between cluster heads and base station is done directly in the LEACH. Although this specification of LEACH avoids energy hole problem but causes the energy of cluster heads that are far from the base station to be discharged faster than others. HEED (Younis & Fahmy 2004) is another well-known clustering based routing algorithms in WSN. Cluster head selection algorithm is based on a relationship between remaining energy and reference energy in HEED. In this paper we propose the EBHCR protocol that is a new cluster head selection algorithm. In addition, data transmitting between ordinary nodes and cluster head and also between cluster heads and base station is multi hop in most rounds in proposed protocol. The rest of the paper is organized as follow: section 2, describes the related works. Section 3 explores the proposed algorithm with detail. Section 4 explain the simulation parameters and result analysis. Final section contains the conclusion. Related Works: As mentioned above in previous papers many protocols for clustering have been suggested. In this section we explain the some selected clustering protocols. MCBT (Shin et al., 2009) proposes a distributed algorithm to create a stable backbone by selecting the nodes with higher energy or degree as the cluster heads. LNCA (Xia & Vlajic 2007) introduces a novel clustering algorithm which uses the similarity of sensed data as an important factor in cluster formation. HEED periodically selects cluster-heads. In this protocol, cluster-head selection is primarily based on the residual energy of each node. HEED is also considered intra-cluster communication cost as a secondary clustering parameter. Clustering algorithm in HEED is done in three phases that is described in details. In first phase, each node identifies its neighbors in cluster range and then computes the required energy for communication with them and also calculates the primary cluster head selection probability by equation (1). CH prob = C prob * (E r / E m Corresponding Author: Roghaiyeh Gachpaz Hamed Computer Engineering Department Islami Azad University, Qazvin, Iran E-mail: ramisahamed@gmail.com 1
Where C prob is the initial percentage of cluster-heads among all n nodes, E r is the estimated current residual energy in the node and E m is a reference maximum energy. In the second phase each of nodes sends an advertising message to the other nodes of its cluster and also receives the same message from other candidates. This advertisement message contains the obtained values from the first phase. Each of these candidates cancel candidacy if its cost is more than the other candidates. In the last phase, non-cluster head nodes attempt to select a cluster head and join to it. LEACH contains four phases that are the advertisements, cluster formation, scheduler creation and data transmission. In first phase, nodes compete to each other for election as cluster head, so that all nodes produce a random number between 0 and 1, then produced number be compared with threshold value which is achieved from (2). If produced number is smaller than T(n), then the node is selected as a cluster head. T( n ) 1 p( 0 p r mod( 1 p )) ng others In this formula, p is percent of cluster heads per all nodes, r is current round and G is set of nodes that are not selected as cluster head in 1/p of last rounds. As can be inferenced, cluster heads election in LEACH is random operation. In the second phase, nodes join their near cluster head and forms the clusters. In the next phase, cluster heads create the scheduler such as TDMA. In the last phase, all nodes transmit the data to their cluster heads based on the created scheduler and cluster heads also aggregate the data before directly sending to the base station. This action will cause unbalanced energy consumption among the cluster heads and therefore lifetime of some nodes is much less than the lifetime of entire network. Proposed Protocol: As mentioned above, EBHCR is the new clustering algorithm.. First, we explain the assumptions used in EBHCR. Then we will describe the EBHCR protocol. Definition: A cluster head is a high level one if its distance to base station is less than the distance of sender cluster head to the base station. An ordinary node is a high level one if its distance to cluster head is less than the distance of sender ordinary node to cluster head. Assumptions: We assume that all nodes are static and also are randomly distributed. At the beginning, the initial energy of all nodes is considered to be equal. All nodes in the network are aware of their location (by GPS or other positioning schemes suchas (Shi et al., 2009) and also are able to control their energy consumption. Let us assume that the base station is aware of remaining energy of cluster heads. Cluster heads are aware of their remaining energy and also of the remained energy of their high level cluster heads. Moreover, ordinary nodes are aware from their remaining energy and also of the remained energy of their high level ordinary nodes. Energy Consumption Model: In EBHCR, energy model is obtained from (Heinzelman et al., 2000) that use both of the open spaces (energy dissipation d 2 ) and multi path (energy dissipation d 4 ) channels by taking amount the distance between the transmitter and receiver. So energy consumption for transmitting a packet of l bits in distance d is given by (3). le l d 2, d d elec fs 0 E ( l,d ) Tx le l d 4, d d elec mp 0 In here d 0 is the distance threshold value which is obtained by (4), E elec is required energy for activating the electronic circuits. ε fs and ε mp is required energy for amplification of transmitted signals to transmit a one bit in open space and multi path models, respectively. fs d. 0 mp Energy consumption to receive a packet of l bits is calculated according to (5). 1377
E ( l Rx ) le elec EBHCR Protocol: EBHCR has four phases that are cluster head announcement, cluster formation, schedule creation and data transmission. The cluster head announcement phase which is cluster head selection phase is almost like a cluster head selection algorithm in HEED but the difference is that in the beginning, all nodes calculates the probability of cluster head selection by (6) and then follows from the operations in the HEED. Competence _Value CH E N ( r ) ( E N m elt n D ) ( ref D D CH _ BS ref ). In here and influence coefficients of energy, number of neighbor nodes and distances, respectively. E r is remaining energy of sensor node and E m is initial energy of sensor. N elt is the number of neighbor nodes of desired node that their remaining energy is less than a threshold. N n is the number of nodes in network. D ref and D CH_BS are the longest distance to the base station (same as network diameter) and the distance between desired node and base station, respectively. In cluster formation phase, all ordinary nodes calculate the merit value of the cluster heads that are on their radio transmission range by (7). Then they join to the cluster head that its merit value is greater than the others. Merit_Value CH E E n CH. ( D D ) 2 n _ CH CH _ BS In here E n is remaining energy of node and E CH is remaining energy of cluster head. D n_ch is the distance between ordinary node and the cluster head. D CH_BS is the distance between cluster head and base station. Schedule creation phase is the same as LEACH in EBHCR. Main difference is in data transmission from cluster heads to the base station. This operation is done directly in the LEACH, but in the EBHCR it is hierarchical. EBHCR considers the distance between ordinary nodes and cluster head and also distance between cluster heads and base station in multi hop and hence can solve the unbalanced energy consumption problem. For this reason, EBHCR uses energy balancing and dynamic hierarchical routing algorithm namely EBDHR which is proposed by (Rasouli Heikalabad et al., 2010) in data transmitting phase. Authors has used EBDHR in (Rasouli Heikalabad et al., 2010) only for inter-cluster routing. But we use EBDHR for inter-cluster and intracluster routing. Simulation and Result Analysis: EBHCR and LEACH are simulated with GCC and the simulation is repeated for many times with different simulation areas and number of nodes to achieve the reliable results for the proposed protocol. Simulation parameters are provided in Table I and obtained results are shown below. In simulation, we assume the values of parameters and are equal. Table 1: Parameters `````````` Simulation area Base station location Radio transmission range Number of nodes Initial energy E elec ε fs ε mp d 0 E DA Data packet size Beacon packet size The desired percentage to become a cluster-head Value 200 meters 200 meters (0, 0)m 40 m 200 3J 50 nj/bit 10 pj/bit/m 2 0.0013 pj/bit/m 4 87 m 5 nj/bit/signal 8192 bits 24 bits 5% 1378
Fig. 3 shows the residual energy of whole network per round for EBHCR and L`1EACH. As it can be seen, obtained values for EBHCR is better than LEACH. Fig. 4 shows the number of dead nodes per round in EBHCR and LEACH. As it can be seen, proposed protocol has a performance better than LEACH protocol so that in EBHCR, the times of first node dies (FND) and half nodes alive (HNA) and last node dies (LND) are optimized about 47%, 35% and 53%, respectively compared to the LEACH. Fig. 1: Residual energy of whole network per round with default parameters. Fig. 2: Number of dead nodes per round with default parameters. Conclusion: This work proposes a hierarchical routing protocol based on new clustering algorithm for wireless sensor networks named EBHCR. The main contribution of this algorithm is to select a best node in terms of energy, number of neighbors and distance as a cluster head. Also the routing algorithm used in proposed protocol is multi hop in most rounds so that can balance the energy consumption among nodes. Simulation results show that the EBHCR prolongs the network lifetime about 45% in comparison to the LEACH. REFERENCES Heinzelman, W.B., A.P. Chandrakasan, and H. Balakrishnan, 2000. Energy-Efficient Communication Protocol for Wireless Microsensor Networks, Proceedings of the Hawaii International Conference on System Sciences. Heinzelman, W.B., A.P. Chandrakasan, and H. Balakrishnan, 2002. An Application- Specific Protocol Architecture for Wireless Microsensor Networks, IEEE Transactions on Wireless Communications, 1: 4. 1379
Hou, T.C. and T.J. Tsai, 2001. Distributed Clustering for Multimedia Support in Mobile Multihop Ad Hoc Network, IEICE Transactions on Communications, E84-B(4): 760-770. Jiang, Q. and D. Manivannan, 2004. Routing Protocols for Sensor Networks," IEEE Consumer Communications and Networking Conference, pp: 93-98. Lin, C.R., M. Gerla, 1997. Adaptive clustering for mobile wireless network, IEEE JOURNAL on Selected Areas in Communications, 15(7): 1265-1275. Rasouli S., Heikalabad, A. Habibizad Navin, M. K. Mirnia, S. Ebadi, and M. Golesorkhtabar, 2010. EBDHR: Energy Balancing and Dynamic Hierarchical Routing algorithm for wireless sensor networks, IEICE Electron. Express, 7(15): 1112-1118. Ryu, J.H.S., H. Song and D.H. Cho, 2001. Energy-Conserving Clustering Scheme for Multicasting in Two-tier Mobile Ad- Hoc Networks, IEEE Electronic Letters, 37(20): 1253-1255. Shi Q., H. Huo, T. Fang and D. Li, 2009. A 3D Node Localization Scheme for Wireless Sensor Networks, IEICE Electron. Express, 6(3): 167-172. Shin, I., M. Kim, M.W. Mutka, H. Choo and T.J. Lee, 2009. MCBT: Multi-Hop Cluster Based Stable Backbone Trees for Data Collection and Dissemination in WSNs, MDPI Sensors, 9: 6028-6045. Soro, S., W.B. Heinzelman, 2005. Prolonging the Lifetime of Wireless Sensor Networks via Unequal Clustering," Proceedings of the International Workshop on Algorithms for Wireless, Mobile, Ad Hoc and Sensor Networks. Xia, D., N. Vlajic, 2007. Near-optimal Node Clustering in Wireless Sensor Network for Environment Monitoring, 21st International Conference on Advanced Networking and Applications, pp: 632-641. Younis, O. and S. Fahmy, 2004. HEED: A Hybrid, Energy-Efficient, Distributed Clustering Approach for Ad Hoc Sensor Networks, IEEE Transactions on Mobile Computing, 3(4): 366-379. Younis, O., S. Fahmy, 2004. Distributed Clustering in Ad-hoc Sensor Networks: A Hybrid, Energy- Efficient Approach, Proceedings of IEEE INFOCOM, 1: 629-640. 1380