Research On Nor-uniform Clustering Routing Protocol In Wireless Sensor Networks

Wireless sensor networks is a kind of network system which combines a large number of cheap micro-sensor nodes within the detection area by way of wireless communication, nodes in the network through the cooperation way monitor information、aware and collect various perceived information in real time within the region, through process these information we can obtain detailed and accurate information eventually. A new generation of information technology is changing from the Internet to the things (Network of Things), from the information communication between persons to the direction of the exchange of information between people with things and content with content. The Wireless Sensor Network is an important part of Things perception layer and the network layer, with a wide range of applications, it can play an irreplaceable role in national defense, environmental monitoring, health care, transportation management, building status monitoring, etc. Since the energy in wireless sensor network is limited, the design of this type of network routing protocol should be as simple as possible in order to reduce the exchange of information between the nodes, reducing the amount of data stored in the node.Routing protocol is the core technology of wireless sensor network layer, whose main task is responsible for forwarding data packets from the source node to the destination node through the network, when forwarding data packets node should discover a small delay path to avoid network congestion and balance network energy consumption. For different application environments, WSN routing protocols can be divided into plane routing protocol, routing protocol and geographic clustering routing protocols and other programs. In the design of routing protocol, it will face a major challenge, namely how to complete perception, communication and control functions in the energy resource nodes, computing power, storage space and communication capabilities are highly constrained conditions. To this end, the main goals of designing wireless sensor network routing protocol are:build energy-efficient path, forming a reliable data transfer mechanism, to maximize the network life cycle.Clustering routing protocol as a hierarchical routing protocol design, we can use it to divide range large and more nodes network into a plurality of small-scale clusters in accordance with specified rules,so the hierarchical of network will be more clearly, more scalable. Since most non-cluster head node can close the communication module in a long period of time, effectively extending the lifetime of the network. But for traditional uniform clustering routing protocol, since the multi-hop between clusters, cluster head near the base station undertakes more forwarding tasks, thus consuming more energy, causing the problem of uneven energy consumption, researchers call this phenomenon as "hot zone" problem. For solving this problem the non-uniform clustering approach came into being. But still there are some problems in the existing non-uniform routing protocols:for example, randomly selected candidate cluster heads, did not consider the remaining energy of nodes. How to improve the success rate of data transmission between clusters. When selecting the next-hop cluster heads based on cost function, how to adjust the parameters of weight in the cost function according to the current changes in the network.For the issues raised above, we propose adaptive non-uniform clustering routing protocol based on entropy weight coefficient method, the proposed solution in this agreement as follows:(1) For the selection of the candidate cluster head by random, this paper sets the threshold formula for each node based on the residual energy of the current network nodes. By taking into account the residual energy in threshold such that the less energy generated by the node has smaller threshold, which has a smaller probability of becoming a candidate cluster head node.(2) In order to improve the success rate of inter-cluster data transfer, we use Bayesian to estimate the link of each cluster head node after each round cluster formed. This can improve the success rate of data transfer between clusters, ensure the integrity of the information collected by the network and pave the way for the selection of a better link state next node.(3) In order to adjust the parameters of weight in the cost function according to the current changes in the network, we use entropy weight coefficient method in formula. We can adaptively select the next hop by accurate and objective determination of the weight ratio of four factors (cluster head residual energy, cluster head distance to the base station, the number of members of the cluster head, the link quality of node to the next hop cluster head) in the share of multi-hop communication process. When node energy is limited, we can as much as possible balance network energy consumption and prolong network lifetime.Finally, we use OMNeT ++ simulation platform to simulate ANCEC algorithms and comparative protocol LEACH, EEUC and DEBUC, the experiment results show that the proposed agreement ANCEC is better than others on data transmission and energy efficiency, it is more stability, efficiency,using ANCEC protocol can effectively extend network lifetime, balance network load, improve the success rate of data transmission.