Research On Clustering Routing Protocol For Wireless Sensor Networks Based On K - Means Clustering And

The sensor technology, computer technology and communication technology are the basis of modern information technology, which have been well applied into the field of information awareness, data processing and transmission and so on. With the users demand increasing, wireless sensor networks becomes a new tool for perceived physical world, and is "Nerve Endings" of internet of things. The research of wireless sensor networks in various countries and regions has paid much attention, and invested a lot of money into the research.Wireless sensor networks is composed of many sensor nodes, which are deployed in the monitoring area. The scale of wireless sensor networks is huge, and the nodes are very limited in energy, information storage and transmission distance. So how to improve the energy utilization to prolong the life time of the nodes, and then prolong the life cycle of the whole networks which has become a key problem for the scholars in the world. Many studies show that the routing protocol based on clustering can reduce the energy consumption of the nodes effectively, manage easily, and the scalability of the network can also be improved.This paper studies the clustering routing protocol, the main work is as follows:(1)Analyzing the compositions and structures characteristics of the wireless sensor networks, introducing the design characteristics of the wireless sensor networks routing protocol and requirements, analyzed some typical routing protocols. And then some examples are given to illustrate the current typical routing protocols, analysis their advantages and disadvantages, pointing out the research focus and the problems to be solved.(2) Introducing the K-means algorithm and particle swarm algorithm, analyzing the application environment and features in detail, including their application environments and characteristics, and then introducing some improvements of several typical methods.(3)Aiming at the issues that wireless sensor networks nodes energy is limited and difficult to supplement, we propose a new routing protocol named K-DENC, which using K-means clustering algorithm to clustering, so as to realize the networks clustering is more compact and reliable, reduce energy consumption of nodes because of the problem that communication distance is longer.(4) After completing clustering, we introduce the evaluation functions, evaluation functions not only considering the residual energy of the nodes but also the nodes position, using the improved particle swarm algorithm to select 2 clusters heads which include the master clusters heads and the vice clusters heads. Using 2 clusters heads model can further reduce the energy consumption and balance the energy consumption of the whole networks. Compared to the classic LEACH protocol, experiment results show that K-DENC protocol can balance energy consumption better, and the nodes energy utilization rate is higher, which can effectively prolong the network life cycle.