The Research Of An Energy-aware Multi-sink Routing Protocol Based On Grid Clustering

As an integral part of the wireless sensor network (WSNs) research, routing protocolshave drawn more and more attention of the scholars with the continuous development ofwireless sensor network. So far, many routing protocols have been put forward insuccession, although different routing protocols make network data transmits in differentways, excellent protocols have a lot in common, such as high efficiency, less delay,low-power, etc.In traditional single sink network, sensor nodes which are around the sink node willshoulder a large amount of data forwarding tasks, so they will be dead, which affects theperformance of the network. Single sink network has some drawbacks, for example,unbalanced network energy consumption and sudden network collapse as a result of thefailure of sink node, with the demanding technical requirements of a variety of applications,single sink network architecture cannot satisfy actual needs. In view of these, multi-sinknetwork become more and more popular for researchers. On the basis of analyzing variousrouting protocols and with the purpose of optimizing the network topology and prolongingnetwork life cycle, this thesis proposes a new multiple sink routing protocol-anenergy-aware multi-sink routing protocol based on grid clustering (EMGC protocol).EMGC protocol divides the network into small squares and each small grid is a cluster.Cluster heads need to send and receive data constantly and heavily, so the energyconsumption of them is much larger than the ordinary nodes’. As a result, residual energy isthe main consideration for the selection of cluster heads in most of the routing protocols.But it is not the only factor in selecting cluster heads, to guarantee the most suitable nodeswill be chosen, EMGC protocol not only think about nodes’ residual energy, but also putnodes’ location into consideration. Also, EMGC protocol uses the periodic rotationmechanism, cluster heads will be re-chosen in the purpose of appointing the more "highquality" nodes. Because of the grid clustering, the distance between two nodes inside acluster is short so these nodes inside a cluster send data to their corresponding cluster headdirectly with the method of single hop to reduce energy consumption. The data transition ofcluster heads adopts energy level mechanism, which means, On the basis of the energy levelof each node, to calculate the energy levels of each possible path, and finally, single out thepath with the largest energy level to transmit data. The nodes taking more energy willshoulder more data forwarding tasks, and then the energy consumption of every member inthe network can be balanced, so that the life cycle of network can be extended as much aspossible. The two main innovations of this thesis are as follows:1. This dissertation introduces the idea of grid clustering, and takes location andresidual energy of nodes into account when choosing cluster heads, thus ensuring the moreappropriate nodes act as cluster heads in order to enhance the balanced of energyconsumption in the network.2. This dissertation adopts energy level model for cluster heads. It keeps the packetstransmitting along the direction of sink nodes, and avoids roundabout transmission path atthe same time. Single out the best path through calculating energy levels of nodes and pathsto make sure that nodes which carry more energy transmit more data, and further enhancethe equilibrium of energy to prolong the life cycle of the network.Finally, the comprehensive performance of EMGC protocol is evaluated in this thesis.The simulations on topology and average nodes energy of EMGC protocol are given byusing the matlab simulation software, also, the comparison between EMGC protocol andELBR protocol is made. The results of these simulation show that EMGC protocol has moresuperior performance than ELBR protocol in the performance of network life cycle, averageresidual energy of sensor nodes: place the same number of sink nodes in the network,network life cycle of EMGC protocol increased by12%than ELBR at least; Place the samenumber of sensor nodes in the network, network life cycle of EMGC protocol increased by15%than ELBR at least; place the same number of sink nodes and sensor nodes and thenetwork running3000rounds, the average residual energy of the network of EMGCprotocol is about1.167times of that of ELBR protocol.