Research On Clustering Algorithm In Heterogeneous Wireless Sensor Networks

Wireless sensor networks (WSN) are currently of concern at home and abroad,the knowledge of which is highly integrated multi-disciplinary involving a high degree of cross-frontier hot research field.It combines sensor technology , information technology and network communication technology.The node not only have perception information and data communications capability, but also have the function of the router. WSN has broad application prospects in the future. Since wireless sensor nodes carry limited energy, the energy efficiency of wireless sensor networks becomes the primary issue of designing WSN routing algorithm. Now it is widely recognized that clustering routing protocol for WSN has better energy-saving features. The traditional clustering routing algorithm for WSN are mostly based on homogeneous networks, but in fact heterogeneous networks applications is much closer to the actual situation.This paper focus on research of clustering routing protocol, and to design an improved algorithm for heterogeneous networks.Firstly this paper describes the general Development of WSN, the concept of wireless routing protocol, compares typical WSN clustering routing protocols. Then we introduce the network model for WSN, the basic principle of clustering algorithm, and analyze the design of clustering algorithm in the heterogeneous network.We study the basic knowledge of the typical clustering routing protocol LEACH, SEP and DEEC,the performance improvements brought by SEP,DEEC compared to LEACH. We also point out shortcomings of SEP, DEEC.We propose the improved algorithm DEEC-k based on DEEC for the limitations of SEP and DEEC.This algorithm introduces a weighted factor k.Adjusting the value of k we can optimize the election probability of a advanced node or an normal one to be a cluster head and this can prolong the lifetime of the network.We simulate DEEC-k algorithm with MATLAB at the case in the two-level heterogeneous networks.We study the relationship between heterogeneous parameters? , m and the network lifetime,and the optimal value of k under specific ? , m .We also analyze the relationship between k and ? , m Qualitatively.Simulational results show that the DEEC-k algorithm provides longer lifetime and higher throughput than the current important clustering protocols in heterogeneous environments.