Multi Data Fusion Model Based on Information Entropy in Sensor Network

Abstract

Data fusion is one of the research hotspots in the field of wireless sensor network. In allusion to the problem of excessive energy consumption of existing fusion methods, a data fusion scheme with minimized energy consumption is proposed in this article on the basis of the compressed sensing theory. Firstly, the influence of different fusion modes on data collection performance is analyzed; then, the influence of route and mixed CS fusion on energy optimization is considered in order to model the data fusion problem into the mixed integer programming problem with minimized energy consumption, and meanwhile a greedy algorithm based on fusion node increase is proposed to solve the above problem, wherein all nodes in the network are divided into fusion nodes used for CS encoding and forwarding nodes used for directly forwarding their own data and the received data; finally, a data fusion tree with minimized energy consumption is obtained. The result of the simulation experiment shows that the scheme proposed in this article is effective and superior to the traditional methods in the aspects of data recovery accuracy, network life cycle, etc.