The Research Of Clustering Topology Algorithm Based On Gradient And Swarm Intelligence Algorithm In WSN

Wireless sensor network is a frontier research field involving multiple disciplines, which has a very broad application prospect, and has been paid extensive attentions. Sensor nodes are often small, and generally work in the harsh environment so that their energy and computing ability is limited. As a result, how to balance the energy consumption and prolong the lifetime occupies a very large weight in the routing design of WSN. With the development of technology, the routing protocol of WSN is gradually inclined to the layered protocol from the planar protocol, showing a trend of hierarchy. As the basis of clustering protocol, clustering topology control algorithm has high energy efficiency and scalability, and has been widely studied and applied.In this paper, we firstly introduced a energy-aware control protocol based on gradient(ETBG). Using the communication radius of nodes, this algorithm divides the whole monitoring area into different gradients, which effectively reduced the height of the cluster tree and the data transmission delay. But the cluster heads are often in biased positions, and clusters are not compact. Furthermore, it doesn’t fully consider the energy and location of gateways, so that some gateways become the weak point of the network.Aiming at the problems existing in ETBG algorithm, this paper proposed a clustering algorithm based on gradient and swarm intelligence algorithm(GSIA). To improve the compactness of clusters, a double cluster head model is established in the cluster, and particle swarm algorithm, which combines the energy of cluster head and the average distance between cluster head and members to construct fitness function,is used to search the optimal nodes to be main and vice cluster head. To optimize the gateways, a path evaluation function and a new pheromone updating model are construct. Using an improved variant ant colony algorithm, paths for each cluster head were established, so that to form a relatively strong cluster tree. Through simulation, we analysis the performance of this algorithm.Furthermore, a new cluster tree optimization scheme is proposed to improve thealgorithm’s stability and reduce the complexity of parameters. It uses particle swarm optimization algorithm, which combines nodes’ location with energy, to search for the best relay nodes for each cluster head and gateway, in order to build a strong cluster tree. The simulation results show that the optimization scheme can effectively improve the stability of the algorithm and increase the network’s lifetime. Finally,illustrate the network maintenance and updating method of GSIA algorithm, in order to improve the adaptability.