Mobile Target Positioning Using Refining Distance Measurements with Inaccurate Anchor Nodes in Chain-Type Wireless Sensor Networks

摘要

As a class of long and narrow structures widely exist such as the river, road, mine tunnel, pipe, chain-type wireless sensor networks (CWSN) can be applied to monitor these environments. The accurate position estimation is a key technology for the mobile target in CWSN. This paper proposes an innovative positioning method to estimate the position of mobile target. Firstly, wireless signals can be affected by measurement noises, coordinate errors of anchor nodes, and chain scene structure. Kernel canonical correlation analysis is applied to analyze the correlation coefficients of these nonlinear wireless signal sets. Secondly, we search out two maximum correlative sets of wireless signals and integrate them into a set of optimal wireless signals. Thirdly, the uncertainty coordinate of anchor node is modeled and the position of mobile target is estimated under measurement and geometry constraints. Furthermore, we simulate the proposed method for mobile target, in comparison with the weighted least squares (WLS) and CHAN methods. Estimation results indicate that the proposed method can refine distance measurement accuracy and perform better positioning performance than WLS and CHAN methods, when we vary the conditions of TDOA/AOA measurement errors, anchor nodes coordinate errors, and anchor nodes spacing distance. Finally, the actual positioning experiments are implemented in a corridor, which show that the practical estimation results are similar to the simulation results.