The Study Of Improved Blind Processing System Based On Wireless Sensor Networks

Wireless sensor networks(WSN) which have the characteristics of low cost, flexible settings and strong self-organization, so the applications and development prospects of WSN are huge. As sensor’s energy is finite and it cannot be charged or replaced in most cases, energy consumption has become one of the main design principles of wireless sensor networks. Blind equalization technique does not use training sequences, so it becomes the first choice of wireless sensor network signal detection technique. In recent years, this team has been devoted to the research of blind equalization technique for Wireless Sensor Networks. Based on preliminary research, the main innovations of this paper are as follows:(1) Based on virtual MIMO WSN system model, in order to improve the performance of the model system, this paper proposes the chaotic positive feedback Hopfield neural network blind detection algorithm and extends it to the MIMO transmission system in outside cluster. This paper utilizes a new second-order statistics blind detection algorithm to recover the sequence of reference sensor in intra-cluster. Compared with the previous model system, simulation results demonstrate that the improved the intra-cluster system has higher speed and improved cluster system has better performance. The overall performance of the improved system is better.(2) On the basis of this preliminary model of this team, this paper analyzes the limitations of the model. In order to improve the robustness, transmission efficiency and reliability of WSN systems, this paper constructs a multi-vMIMO hierarchical WSN blind detection model. In this model, the sensors are clustered through different areas or functions. The common sensors in the first layer and multi-cluster heads in the second layer constitute the first-level virtual MIMO transmission system, and a plurality of cluster head and Sink in the third layer constitute the second-level virtual MIMO transmission system.(3) In multi-vMIMO hierarchical WSN blind detection system, intra-cluster sensors’ sequences have correlation, so this paper introduces literature blind algorithm to solve the first-level virtual MIMO system signal detection problem. By improving literature algorithm, so the scope of application of the algorithm is not limited to signals’ correlation. This paper utilizes this improved algorithm in the first-level virtual MIMO system. Simulation results indicate the feasibility of the blind detection system.