An Energy-Efficient Hierarchical Routing for Wireless Sensor Networks

Volume 2 Issue 9, 213, ISSN-2319-756 (Online) An Energy-Efficient Hierarchical Routing for Wireless Sensor Networks Nishi Sharma Rajasthan Technical University Kota, India Abstract: The popularity of Wireless Sensor Networks (WSN) has increased tremendously in recent time due to growth in Micro- Electro-Mechanical Systems (MEMS) & wireless communication technology. The main purpose of such networks is to collect information from the environments and deliver the same to the applications to determine characteristics of the environment or detect an event. As sensor nodes depend on batteries, they have limited amount of energy. One of the major issues in wireless sensor networks is to develop an energy-efficient routing protocol which has a significant impact on the overall lifetime of the sensor network. Many energy efficient routing protocols have been proposed to solve this problem and increase the lifetime of the network. This paper proposes a clustering technique which is based on LEACH protocol. In this, clusters are formed based on their geographical location & cluster heads are sted on the basis of highest residual energy within the cluster as well as minimum distance to the base station from the cluster heads. The network lifetime for various levels of hierarchical routing techniques is compared. There are two outcomes from the implemented protocol. These are prolonged network lifetime & increased mean residual energy with the increase in number of clusters in the wireless sensor network. Keywords: Wireless Sensor Networks, Energy efficiency, Hierarchical routing, Improvement on LEACH protocol, Network lifetime. 1. INTRODUCTION The emerging field of wireless sensor networks combines sensing, computation, and communication tasks into a single tiny device. A wireless sensor network can be generally described as a network of nodes that cooperatively sense and may control the environment enabling interaction between persons or computers and the surrounding environment. It is composed of a large number of sensor nodes that are densely deployed either inside the event area or very close to it. The position of sensor nodes needs not to be fixed. This allows random deployment of sensor networks in inaccessible terrains or disaster relief operations. Sensor nodes are fitted with an onboard processor. Instead of sending the raw data to the other nodes, they use their processing abilities to locally carry out simple computations and transmit only the required and partially processed data [1]. A greater number of sensors allows for sensing over larger geographical regions with greater accuracy. Sensor nodes coordinate among themselves to produce high-quality information about the physical environment. Each of these scattered sensor nodes has the capability to collect and route data either to other sensors or back to an external base station(s). A base-station should be capable of connecting the sensor network to an existing communications infrastructure or to the Internet where a user can have access to the reported data [2]. The sensor nodes depend on batteries as their energy sources which means they are constrained in terms of energy. In order to achieve longer network lifetime, energy efficient routing protocols are required. The Routing protocols are broadly divided as Flat, Hierarchical & Location based routing [2]. The hierarchical or clustering technique is considered as a good method to minimize energy consumption in WSN. It is mainly used by the sensor nodes, in which sensors send information to only the cluster heads and then the cluster head transmits the aggregated information to the base stations or the sink. In this, the nodes are often grouped together into disjoint and mostly non-overlapping clusters to minimize communication latency and improve energy efficiency. Leader of every cluster is often referred to as the cluster-head. A cluster-head may be ted by the sensors in a cluster or pre-assigned by the network designer. A cluster-head may also be just one of the sensors or a node that is richer in resources. Cluster-heads may form a second tier network or may just ship the data to the base-station. The advantage of this scheme is that it reduces energy usage of each node and communication cost. In this technique [3], nodes perform different tasks in WSNs and typically are organized into lots of clusters according to specific requirements or metrics. In general, nodes having higher energy act as cluster head and perform the task of data processing and information transmission, while nodes with low energy act as member nodes and perform the task of sensing the information. The typical clustering routings protocols in WSNs include LEACH, PEGASIS, TEEN, APTEEN and HEED etc for wireless sensor network. The organization of the paper is as follows. Section I gives introduction in brief, section II explains about LEACH protocol, section III describes the Sensor Network Model. Section IV gives proposed protocol. Section V gives Simulation results of non-hierarchical & hierarchical routing protocols using MATLAB. Section VII concludes the paper. 2. DESCRIPTION OF LEACH PROTOCOL Hierarchical based routing protocols provide data aggregation, scalability, less load, less energy consumption and more robustness, more fault tolerance as compare to flat based routing protocols [3]. Heinzelman et al. proposed Low-Energy Adaptive Clustering (LEACH) [4] for efficient routing of data in wireless sensor networks. LEACH is a selforganizing, adaptive clustering protocol that uses randomization to distribute the energy load evenly among the sensors in the network. In LEACH, the nodes are organized themselves into clusters, with one node behaving as the cluster-head. If the cluster heads were chosen a priori and www.ijcat.com 173

Volume 2 Issue 9, 213, ISSN-2319-756 (Online) fixed throughout the system lifetime, they would die quickly, ending the lifetime of all nodes belonging to those clusters. Thus LEACH uses randomized rotation of the high-energy cluster-head position such that it rotates among the various sensors so that the battery of a single sensor would not drain quickly. LEACH also performs data fusion to compress the amount of data being sent from the clusters to the base station, thereby reducing energy dissipation and enhancing system lifetime. Sensor nodes t themselves as cluster-heads at any given time with a certain probability. These cluster head nodes advertise their status to the other nodes in the network. Each sensor node determines the cluster to which it wants to belong by choosing the cluster-head that requires the minimum communication energy for transmission. Once clusters are formed, each cluster-head broadcasts a TDMA schedule and assign each node a time slot in which it can transmit the sensed data. The non-cluster head nodes sense the data and send it to their cluster-head according to the TDMA schedule. Once the cluster-head collects all the data from the nodes in its cluster, they aggregate the data and then transmit the compressed data to the base station. Since the base station is far away in the scenario we are examining, this is a high energy transmission. The main drawbacks of LEACH protocol are: Cluster heads are sted randomly in LEACH; it is possible that nodes with less energy would be chosen, which could lead to these nodes die too fast. In addition, because in LEACH protocol cluster heads communicate with base stations in single-hop manner, it is energy consuming and its expandability is limited so that it could not adapt to large network. Both of these shortcomings will be overcome by the proposed protocol. 3. SENSOR NETWORK MODEL There has been a significant amount of research in the area of low-energy radios. In this work, we assume a simple model where the transmitter dissipates energy to run the radio tronics and the power amplifier and the receiver dissipates energy to run the radio tronics [4, 5] as shown in Figure 1. The power attenuation is dependent on the distance between the transmitter and receiver. The propagation loss will be inversely proportional to d 2 for relatively short distances, whereas it will be inversely proportional to d 4 for longer distances. Thus, to transmit a k-bit message to distance d, the radio expends: E E TX TX ( k, d) k * E E ( k, d) k * E ( k, d) k * E amp k * * d k * fs mp * d and to receive this message, the radio expends: ERX ( k) E * k Here E TX (k, d) - energy dissipated per bit at transmitter E RX (k, d) - energy dissipated per bit at receiver 2 4 if d d if d d E - energy required while transmitting or receiving one bit of data E amp - amplifier coefficient d - distance between a sensor node and its respective cluster head or between a cluster head to another cluster head nearer to the base station or between cluster head and base station. k bit packet Transmit Electronics Tx Amplifier Figure 1: First order radio model 4. PROPOSED ROUTING ALGORITHM This paper introduces an Improvement to LEACH protocol which increases the lifetime of the sensor network as well as mean residual energy of the network. In the proposed protocol, clusters are formed geographically. The sensing area is divided into the equal parts, nodes belongs to the same part forms the cluster. In case of non-hierarchical cluster formation, entire sensor area space will be used. But in other cases such as two, three & four cluster formation, sensor area space will be divided into equal areas. The two, three & four clusters formation are otherwise known as first level, second & third level respectively. In case of LEACH this equal area segregation is not used. Once the clusters are formed, cluster-head stion phase starts. In this work, we make use of first order radio model to make some energy calculations. We first calculate the residual energy of each node of each cluster. Then the energy of the nodes within a cluster is compared to each other. In order to do efficient communication, the node having maximum energy in the clusters is sted as cluster-heads. Since the cluster head performs data collection from various sensor nodes within the cluster, data aggregation & data transmission, they lose their energy very quickly. Due to draining activities being constraint on a cluster head; the cluster-head is rotated among the sensor nodes of each cluster at every transmission round. After each transmission the residual energy of the nodes is recalculated and then again comparing the nodes energy and sting the node with maximum energy as the cluster-head. By rotating the cluster heads on the basis of residual energy, we can equally divide the burden of transmission & thereby increase the network lifetime. In this approach we also use minimum distance concept during transmission. We calculate the distance between nodes to cluster heads and cluster heads to other promising cluster heads or to the base station. Then the minimum distance path is sted for the transmission so that less energy is wasted in transmitting data. Once the cluster head with shortest path is sted, they aggregate the data to be transmitted and then transmit it using shortest path. The proposed hierarchical routing algorithm can be summarized using following steps (Figure 2): d Receive Electronics k bit packet www.ijcat.com 174

Volume 2 Issue 9, 213, ISSN-2319-756 (Online) Cluster formation is done by dividing the area into equal parts. Cluster heads are sted from each cluster on the basis of residual energy as well as the shortest distance to the base station. Data aggregation phase which involves the gathering of collected data by the cluster head from the sensor nodes within its cluster. Data transmission phase in which data is transferred from the cluster heads to other CHs or to the base stations. The Cluster Head stion process used in proposed technique can be described as follows:- The initial energy E in (n) of each node is measured. Also, the distance d(n) from each node to the base station or to the next higher level cluster head is calculated. Then we compare the measured distances & st the minimum distance for the transmission in that round. Estimation of the energy required by each node for transmission within the cluster (not to BS or to higher level CH) is carried out using the formula: (E amp *k*d 2 ). The residual energy of each node is given by E ( n) E in in ( n) ( E * k E amp * k * d The cluster head is then sted on the basis of maximum residual energy in that round. After the CH stion is done, the next cluster head stion will take place after the current round is completed. This cluster head stion process is shown in Figure 3. 2 ) Start Read the number of cluster required M=1,2, 3, 4 Sensor node deployment with n=4 Clusters formation based on M Round=x, Count=, i = 1 CH(i) stion based on Estimated Energy as well as shortest distance to the Base station i = i + 1 Is i > M Data Aggregation Data Transmission Count= Count + 1 Is Count< x Display Network Lifetime of the system & Residual energy of the Senor network Stop. Figure 2: Flow chart of Proposed Routing Technique www.ijcat.com 175

Volume 2 Issue 9, 213, ISSN-2319-756 (Online) Est (j)=, j= j+1 First Initialization step (i.e Round = x) in Figure 1 n=4 & j=1 Is node (j) ε M Measure initial Energy E in(node (j)) Calculate the Min{d(node (j))} with respect to base station or next CH { if M!=1} show how this protocol increases network lifetime & residual energy as the number of clusters is increased. Throughout the simulation, 3 sensor nodes are deployed in 3 x 3 regions. The size of sensor data packet is K bytes. We use first order radio model for data reception, aggregation & transmission by the nodes in the sensor networks. are initialized with J energy. Then we compare the results with the network of nodes having initial energy 3J. Results show that network lifetime increases when we increase the initial energy of the nodes. Simulation parameters are shown in table 1. Table 1: Simulation parameters Parameter Value Number of 3 Network size(m 2 ) 3 3 Base station location (, ) Initial Energy J/3J Packet size (Kbytes) E (nj/bit) 5 Calculate estimated residual energy for each node Est (j) E ( node in ( j) ) [ E * k E amp *{ Min( d( node ( j) 2 )} ] E amp (pj/bit) For simplicity, we assume the following: All nodes are homogeneous in nature; Is j < n? j = j + 1 All nodes begins with the same initial energy; Clusters and nodes are static. This means nodes location is fixed throughout the operation; Evaluate Jmax (i)=max(est 1, Est 2, Est 3,Est n) Normal nodes transmit the data directly to their respective cluster heads within a particular cluster; CH(i) = Jmax (i) Predefined algorithm which leads to the second initialization (i = i+1) in Figure 2 Figure 3: Flowchart of Cluster head stion process 5. SIMULATION & PERFORMANCE EVALUATION In this section we present the results of experiments carried out for evaluating the performance of the proposed hierarchical based routing for one, two & three clusters, four. Simulation experiments are carried out in MATLAB. We also Cluster heads use multi-hop routing to relay data to the base station. They can transmit data either to base station or to next promising cluster heads. 5.1. Node Deployment The sensor nodes in the network are formed into clusters of different sizes of one, two, three & four. One indicates a Nonhierarchical formation of cluster whereas two, three & four indicate First, Second & Third level. Figure 4 indicates the non-hierarchical structure of our routing technique. Likewise, Figure 5 and 6 shows the simulation result of two & three cluster formation using proposed hierarchical technique. We also show the simulation result of four clusters formation scenario which indicates Third level hierarchical technique. In this case formation of clusters is done in two ways (Case I & II) shown in Figure 7 & 8. www.ijcat.com 176

Y Co-ordinate Y Co-ordinate Y Co-ordinate Y Co-ordinate Y Co-ordinate Volume 2 Issue 9, 213, ISSN-2319-756 (Online) 3 3 256 27 4 197267 18 264 223 163 22 24 176 63 34 221 26 31 38 263 227 11274 61296 26 4 181 65 9 11 9 186 187 189 11 149159 172 164 111 79 98246 294 153 252 249199 279 293 241 25 157 42 13 135 136 54 57 141 235 217 18 6132 155 232 154 284 71 3 62 35 37 44 59 184 261 193 219 115242 278 129 19266 95 289 23 288 36 183 127 48 5 45 166 228 254 68 12 27 148 3 277 12 52 66 75 118 131 137 142 262 192 287 25 117 15 119 17 243 274 276 285 122 125 171 143 19 21 233 224 257 73 2 251 32 7 94 87 124 168 196 185 161 53 14 144 158 116 96 23 26 25 291 15 18 126 194 198 229 17 121 58 162 167216 83 25 299 239 24 99151 178 22 67 17 165 225 265 33 258 13 92 182 134 16 292 29 23 145 128 147 78 247 51 56 41 72 177 245 253 295 282 21 114 21 91 218 283 244 88 17312 152 212 191 146 24 215 259 248 169 19 43 8986 28 13 29 16 6 213 15 64 297 133 29226 255 8549 231 222 273 5 1 123 237 22 286 77 27593113 5 76 7 46 8 13814 15 11 139 82 39 16 179 188 27 268 271 214 272 281 288 234 174 195 14 156 211 236 298 238 69 81 47 97269 175 1 28 55 5 15 25 3 Figure 4: Non hierarchical cluster formation 3 57 164 187 228 239 224 262 251 53 76145 95 269 276 15 69 18 162165 261 78 8 141 243 299 295 197 19 93 36 114 129 263 268 132 286 232 55 89 283 118272 83 2212 25 47 63 194 51 58 128 223172 21 176 85 16 39142 98 179 174 124 226 234 257 277 94 82 15959 219 8 14 7 26 35 48 2 7915 87 22 254 271 173 266169 88 135 229 215 4 27 144 34 64 126 188 285 16 11 252 23 29429 244 6228 61 143 217 279 222 185 236 12 9 3 9 1 253247 256155 74 21 73 17 33 72 123 147 152 24 167 166 3 15 148 19 49 12 186 16 213 293 242 214 2259 97 154 113 116 138 182 265 27 54 43 23175 221 225 287 248 237 127 14 24 38 65 112 7 199 17 189 291 246 122 275 117 28 125 193 245 258 267 28 281 4 71 25 96 249 45 255 115 183 19625 6611 46 37 185686 211 233 81 171 29 137 24 288 157192 13 133 13 163 198 297 178 235 274 296 264 31 92 1117 6 168 1 5 91 19 216 151 153278 227 238 44 5 77 68 273 292 99 52 3 121 282 15 195 26 29 284 298 41 6 23 184 25 75 67 181 149 11 1813 42 191 2 5 1284 14 134 26 231 139 289 136 131 158161 177 156 241 27 32 146 21 218 119 5 15 25 3 Figure 5: First level hierarchical cluster formation 3 37 124 7 157 18 1 149 117 237 49 228 175 26 165 235 267 246 289 3 83 227 167 229 224 19 24 8 92 43 86 23 25 25 134 214 159 258 257 284 249 291 295 77 29 264 22 36 4 97 11 56 85 34 15 13 12 156 185 283 239 259 173 6 66 57 145 178 148 188 196 55 12 217 266 44 9 128 48 132 216 94 99 139 152 172 81 91 297 245 29 123 162 12 256 262 299 136 59 16 41 42 191 271 176 215 18 197 147 58 6 8 192 22 263 244 253 1 19 154 125 15 32 183 226 236 142 171 261 29 268 19 286 186 114 273 194 276 93 47 67 13 119 28 27 182 232 74 9 15 51 87 11 242 141 195 11 84 75 62 71 126163 14 252 278 285 76 272 213 27 1618 15 135 4 279 17 21 118 238 52 181 45 96 164 168 26 248 219 28 5439 5 112 179 24 247 269 255 25 122 79 2 177 17 138 31 22 254 189 146 296 28 293 64 27 13 174 23 155 221 218 5 72 82 14 223 24 33 129 251 193 95 98 288 265 73 88 3 274 2 275 131 211 5 116 53 198 294 16 38 137121 212 25 225 111 292 151 243 61 199 69 127 89 17 26 68 2317 14 113 187 144 158 241 63 233 115 21 222 35 298 169 287 23 133 65 46 143 153161 184 3 78 21 234 277 281 166 282 5 15 25 3 Figure 7: Third level hierarchical cluster formation 3 182 241 75 153 194 12 169 234 197 227 22 63 72 6 42 93 286 289 215 137 79 294 74 276 59 3146 113 57 13 3 24 29 84 112 82 161 27 192 73255 189 252 187 163 81 16 25 183 23 47 133 24 71 152 222 13548 18 287 143 277 275 262 15 17 284 33 225 87 23 27 186 217 168211 4 96 253 53 13 19358 19 224 88 174 23 26 249 274 271 228 246 13 2 25 259 26 273 278 291 297 298 164 44 89 128 52 19 154 237 213 283 265 49 5 11 282 99 43 41 15 115 142 191 14 18 171 219 244 147 257 3 12238 176 243 184 21 238 266 212 214 61 12 175 25 218 118 268 216 148 32 68 14 156 15 29 272 16 8 121 134 2 85 195 123 292 29 136 1 7 16 98 229 21 232 157 55 139 167 281 27 173 258 239 97 3 4 155 251 24 256 9 226 76 62 35 83 116 296 132 129 25 269 23595 264 288 165 9 146 138 94 245 1736 293 45 37179 91 223 56 66 18 126 172199 114 1 131 162 144 231 261 285 299 11 15886 14122 177 117 22 221 5 64 149 247 54 34 15167 2428 19 77 17 124 21 236 166 254 6 69 28 125 78 92 196 51 12 198 267 28 279 26 145 5 119 15 159 39 233 181 263 28185 65 7 111 127188 248 295 11 178 14 5 15 25 3 3 262 45 27 93 235172 268 266 179 272 294 13 114 227 243183 23 275 95 25 12 34 241 139 213 15 164 16 44 59 162 21 225 242 287 128 297 25 61 238 273 119 191 177 67 152 89 26 195 224 11 94 22 11 154 174 116 284 72 13 69 299 16 236 6 277 186 25 14 125 35 5 127 3 282 8 166 248 27 136 146 13 1548 23 14797226 278 167 9 14 134 192 2533 43 255 291 296 91 138 132 36 52 71 57148 84 181 259 2 4 157 144 133 53 214 171 185 16 51 149 98 194 292 163 184 32 199 26 234 217 122 15 47 156 215 46 37 143 28151 247 121 252 82 27 19 111233 197 24 66 62 18 65 96 219 285 286 88251 221 168 6 182 24 126 99 175 245 187 49 31 86 24158 5 264 17 288 87 41 239 1 9 216 11 293 21 237 246 141 42 68 276 279 222 3 131 212 271 145 258 78 55169 23 92 142 64 269 295 281 263 28 83 15 547 26 63 188 33173 19 265 196 38 81 85 5 19 7 117 137 14 28 113 18 12 211 1 155 176 165244 22 29 256 75123 254 267 283 12 73 159 261 17 4 77 39 18 193 76 79 228 17 21 232 231 298 115 22 29 198 223 118 2 25 58 189 29 74 129124 112 135 153 178 218 229 257 289 274 8 249 161 5 15 25 3 Figure 6: Second level hierarchical cluster formation Figure 8: Third level hierarchical cluster formation 5.2. Network Lifetime One of the most important factors for evaluating the sensor network is network lifetime. The lifetime of the network depends on the lifetime of each sensor node that is a part of that network. Recharging and replacing the nodes batteries is impractical in many environments. In our experiment we define the network lifetime as the time until all the nodes are dead. The network lifetime for non hierarchical, First level, Second level, Third level hierarchical routing is shown in Figure 9,1,11,12, & 13. We observed that the nodes begin to die more quickly in the non-hierarchical routing technique since all nodes in the network send captured data via one randomly sted cluster head per round to the base station. www.ijcat.com 177

Number of Alive Number of Alive Number of Alive Number of Alive Number of Alive Volume 2 Issue 9, 213, ISSN-2319-756 (Online) We observed that the proposed technique offers improvement in network lifetime as we increase the number of clusters in the network. Non-hierarchical technique network completely stopped functioning at an earlier simulation rounds compared to our proposed technique. We observed that Case II of Third level gives better results than Case I. Table 2 displays number of alive nodes with the increase in number of clusters when the initial energy of all the nodes are Joules. We further increase the initial energy of all the nodes to 3 Joules. Results are shown in Table 3. We observed that when we increases the initial energy of all the nodes, network lifetime will be increased. 3 25 15 5 3 25 5 15 25 3 35 4 Figure 11: Network lifetime graph (Second-level hierarchical Routing protocol) 15 3 25 5 5 15 25 3 35 4 15 Figure 9: Network lifetime graph (Non-hierarchical Routing protocol) 5 3 25 5 15 25 3 35 4 Figure 12: Network lifetime graph ( hierarchical Routing protocol Case I) 15 3 25 5 5 15 25 3 35 4 15 Figure 1: Network lifetime graph (First-level hierarchical Routing protocol) 5 5 15 25 3 35 4 Figure 13: Network lifetime graph ( hierarchical Routing protocol Case II) www.ijcat.com 178

Residual Energy in Joules Volume 2 Issue 9, 213, ISSN-2319-756 (Online) Table II: Network Lifetime Comparison (Initial Energy=Joules) Table III: Network Lifetime Comparison (Initial Energy=3Joules) First level Second level First Node Half Last Network Lifetime 4 74 124 12 43 161 191 148 74 243 334 26 alive till last round Non Non First level Second level First Node Half Last Network Lifetime 7 127 197 19 187 375 - - 61 189 - - - 188 alive till last round Third level 65 249 - - 36 Third level 127 - - - 155 91 271 - - 8 Thirdlevel 236 - - - 242 5.3. Mean Residual Energy We evaluate the residual energy of each node for particular rounds of simulation. Simulation results in Figure 14, 15 & 16 show residual energy of nodes in non-hierarchical, first level & second level hierarchical routing technique respectively. Results show that the mean residual energy value of nodes in the proposed hierarchical technique is higher than non hierarchical technique. This implies improved network performance since the nodes has more energy. We also evaluate the mean residual energy of nodes for third level (four cluster formation). Simulation results in Figure 17 & 18 shows increased residual energy as compare non hierarchical, first & second level. Case II performs better than case I in terms of residual energy. The mean value of the residual energy increases as the number of cluster increases. 18 16 14 12 8 6 4 2 Graph Illustrating Residual Energy Vs 5 15 25 3 Figure 14: energy residue in non hierarchical technique after 4 rounds of simulations. www.ijcat.com 179

Residual Energy in Joules Residual Energy in Joules Residual Energy in Joules Residual Energy in Joules Volume 2 Issue 9, 213, ISSN-2319-756 (Online) 18 Graph Illustrating Residual Energy Vs 18 Graph Illustrating Residual Energy Vs 16 16 14 14 12 12 8 8 6 6 4 4 2 2 5 15 25 3 5 15 25 3 Figure 15: residual energy in First-level hierarchical technique after 4 rounds of simulations. Figure 18: residual energy in hierarchical technique after 4 rounds of simulations 18 16 14 Graph Illustrating Residual Energy Vs Simulation results are shown in Table 4 when the initial energy of the nodes are Joules. We also observed that the residual energy of the nodes & network is also increased when we increases the initial energy of the nodes to 3 Joules. These results are shown in Table 5. 12 8 6 4 2 5 15 25 3 Figure 16: residual energy in Second-level hierarchical technique after 4 rounds of simulations. 18 16 14 12 8 6 4 2 Graph Illustrating Residual Energy Vs 5 15 25 3 Figure 17: residual energy in hierarchical technique after 4 rounds of simulations. Table IV: Comparison of Mean & Variance of Residual Energy (Initial Energy= Joules) Mean residual energy (Joules) Variance residual energy (Joules) Non-hierarchical 8.33 8.5 First-level 11.68 12. Second-level 53.65 41.62 67.95 44.24 81.77 49.22 Table V: Comparison of Mean & Variance of residual energy (Initial Energy= Joules) Mean residual energy (Joules) Variance residual energy (Joules) Non-hierarchical 11.31 8.65 First-level 85.61 4.64 Second-level 132.9 66.1 11.31 65.11 178.35 62.47 www.ijcat.com 18

6. CONCLUSION Simulation results show that the proposed hierarchical routing technique gives better results when compared to nonhierarchical technique. We observed that network lifetime & residual energy of the network is increased when we increase the number of clusters. This means that network remains alive for a longer time so more transmission can be done when proposed technique is used. This also shows that when we increase the initial energy of each node, the network lifetime & residual energy will also be increased. 7. REFERENCES [1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, "A survey on sensor networks," Communications Magazine, IEEE, vol. 4, pp. 12-114, 2. [2] Al-Karaki, J. N. and A. E. Kamal, "Routing Techniques in Wireless Sensor Networks: A Survey," IEEE Wireless Communications, 11(6), Dec., 4. [3] Xu-Xun Liu A survey on Clustering Routing Protocols in Wireless Sensor Networks, Sensors 212. [4] W.Heinzelman, A.Chandrakasanand and H. Balakrishnan, "Energy-efficient communication protocol for wireless microsensor networks," in Proceedings of the 33rd Annual Hawaii International Conference on System Sciences, Maui, HI,, pp. 35-314. [5] W.B. Heinzelman, A.P. Chandrakasan, H. Balakrishnan, "Application specific protocol architecture for wireless microsensor networks," Wireless Communications, IEEE, vol. 1, pp. 66-67, october 2. Volume 2 Issue 9, 213, ISSN-2319-756 (Online) www.ijcat.com 181