Research Of Data Fusion Algorithms For Target Location

Wireless sensor networks(Wireless Sensor Networks, WSNs) is a multi-disciplinary research area, involved in several fields such as signal processing, information theory, statistical estimation and inference, and artificial intelligence etc. While wireless sensor networks is facing several challenge about energy-constrained, communication capability and data processing etc. In this paper, it discusses in detail sensor data, noise processing and target tracking problems for target localization and tracking in WSNs. Then on the basis of the structure and characteristics of WSNs, this paper presents several corresponding resolution following the design requirements of energy priority and data centric, and also makes some theoretical research in related fields, laid the foundation for future further research and applications. The main innovations and work include the following aspects:1. For the purpose of improving both of maneuvering and non-maneuvering target tracking performance with mobile sensors deployed in WSNs, it presents an algorithm of mobile sensors dada fusion tracking for WSNs (MSDFT). Combined with the "GATING" technique, it will be able to solve the problem of data association and maneuvering targets by establishing a dynamic structured model-Extended Kalman Filter as adaptive maneuvering compensation, which research an algorithm of mobile sensors tracking multiple targets. Finally, simulation results indicate that the approach can track multiple mobile sensors deployed in WSNs successfully.2. As Data provided by sensors is always affected by some level of uncertainty in the measurements, this paper presents an optimal method of data fusion for multi-sensors based on Bayesian estimation utilizing the data redundancy to reduce this uncertainty and to achieve improved accuracy. The proposed method establishes new evaluation of data fusion relying on combining a Bayesian fusion algorithm with Kalman filter in WSNs. A case is presented to verify the proposed algorithm by estimating the position of mobile robot. Experimental study shows that combining Bayesian fusion algorithm with Kalman filter can help handle the problem of uncertainty and inconsistency of the data in both centralized and decentralized data fusion architectures.3. Successful node localization system is of great importance to design efficient localization algorithms in WSNs. This paper presents a range-free algorithm of localization using Fuzzy logic to achieve more optimum solution. It models edge weights using Fuzzy logic system, which utilizes the current value of the Received Signal Strength Indicator(RSSI) and the Link Quality Indicator(LQI) from the anchor nodes for deciding location of sensor nodes. Simulation results have shown that the proposed algorithm can deal with target localization tracking in WSNs and have demonstrated the effectiveness of the proposed approach.