Environmentally Adaptive Event-Driven Robust Cubature Kalman Filter for RSS-Based Targets Tracking in Mobile Wireless Sensor Network

Abstract

This article presents an environmentally adaptive event-driven robust square root cubature Kalman filter (EAERCKF) for received-signal-strength-based (RSS-based) moving targets tracking in resource-constrained mobile wireless sensor network (MWSN) under the motion modeling and RSS quantization uncertainties. First, an environmentally adaptive event-driven mechanism, scheduling dynamically and approximately the desired amount of anchors regardless of the anchors distribution density near the moving target, is designed to reduce energy consumption, communication rates, and computation complexity of the MWSN system. Second, a robust square root cubature Kalman filter is developed to estimate the motion state of the moving target, where the covariances of the process noise and the measurement noise assumed to be time-varying uniform distributions are generated randomly. Third, the bounds of the uniformly distributed covariances describing the motion modeling and RSS quantization uncertainties are given to improve the stability, accuracy, and consistency of the estimator. Finally, the results of numerical simulations demonstrate that our proposed EAERCKF estimation method has superiority in simplifying the determination of the random noise covariances with satisfactory target tracking performance and approximating the desired number of triggered anchors.