Hybrid Robust Sequential Fusion Estimation for WSN-Assisted Moving-Target Localization With Sensor-Node-Position Uncertainty

Abstract

This article presents a hybrid robust sequential fusion estimation (HRSFE) method for moving-target localization in mobile asynchronous wireless sensor networks (WSNs). The estimation error or inconsistency, caused by the uncertainties of sampling, communication, and sensor-node position, is considered fully and compensated by incorporating the sensor-node position with uncertainty into the state vector as well as introducing an adaptive fading factor into the cubature Kalman filter (CKF) to maintain the estimation consistency. In addition, a QR-decomposition-based square-root form of the cubature strong localization filter (SR-CSLF) is proposed to improve the accuracy and stability of the CSLF. Moreover, an HRSFE method is designed to estimate the augment state of the WSN-assisted moving-target-localization system, and the proposed estimation method combines the merits of both the SR-CSLF and SR-CKF effectively and is able to deal with the uncertainties of sampling, communication, and sensor-node position in a unified framework. Simulations and experiments of a WSN-assisted moving-target-localization system under various uncertainty settings are employed to demonstrate the effectiveness and superiorities of the proposed method.