The Study Of Blind Processing System Based On Clustering Virtual MIMO Wireless Sensor Networks

Wireless Sensor Network(WSN) is the main trend in future which has the characteristics of low cost and strong dynamic topology.It can be widely used.Currently,WSN adopts adaptive equalization technique based on training sequences which need a lot in order to meet the ad-hoc and channel dramatic change network.Low energy consumption is the one of key design principles due to the difficulty of changing the nodes batteries.Training sequence is not needed in blind equalization and detection technology,so it can save energy effectively and has great application value in WSN.Some progress had made in the study of blind equalization and detection technology in WSN.This paper is supported by National Nature Science Foundation(number:61302155).On the basis of our team s early achievement, the main innovations are as follows:(1)In order to obtain more accurate data,based on virtual MIMO wireless sensor network system model,this paper utilizes MIMO-LPA outside cluster and Hopfield Neural Network(HNN) blind detection algorithm intra-cluster.Then,the estimated signals of each node through intra-cluster and outside cluster two layers blind signal processing are gotten from the whole WSN system.(2)On account of the pseudo random of chaos and hyperchaos sequences,this paper proposes precoding technique blind detection scheme intra-cluster based on chaos and hyperchaos.Simulation experiments show that both schemes can improve the performance of blind detection in WSN and show that the scheme based on hyperchaos is better.(3) To further improve the blind processing performance of WSN,this paper proposes Positive Feedback Hopfield Neural Network(PFHNN) hyperchaos precoding wireless sensor network blind processing system due to the PFHNN anti-noise property.Simulation experiments show that PFHNN can improve the blind processing performance while it reduces convergence speed. Then, this paper proposes Double Simoid Positive Feedback Hopfield Neural Network(DS-PFHNN) hyperchaos precoding wireless sensor network blind processing system.Simulation experiments show that DS-PFHNN hyperchaos precoding wireless sensor network blind processing system can improve the performance both in BER and convergence speed.