Tracking moving targets in wireless sensor networks using extended diffusion strategies of distributed Kalman filter

Abstract

Using wireless sensor networks to track the position of a moving object in a 3-D spatial model requires precise information of location and speed of the object, which in turn demands for accuracy in state estimation of distributed Kalman filter. In view of reducing the impacts of noise in the dynamic linear system and achieve optimized state estimate, the current study proposes extended strategies of Kalman filter diffusion based on distributed Kalman filter. Through the proposed strategies, each node communicates merely with its neighbor nodes. The data aggregation is done in a set of neighborhood using instructions of recursive Kalman filter iterations with specific weights. The proposed algorithms provide precise state estimates in a moment as global state estimates using various updates at each step. As a simulation study, we applied the algorithms in a network to track the position and speed of a projectile and compared the results with real world circumstances, using the concept of transient mean square deviations of network as a cost function. The results report improvements over the conventional methods in terms of mean square errors.

Audiovisual Speaker Tracking Using Nonlinear Dynamical Systems With Dynamic Stream Weights

“…To assess the speaker localization and tracking performance, the circular root mean square error (RMSE) (27) was employed as an evaluation metric [46], whereφ k is the estimated azimuth at time-step k, φ k is the corresponding ground-truth azimuth angle, K is the total number of timesteps in one test sequence and k 0 corresponds to the number of frames in the grace period. This metric was calculated individually for each audiovisual test sequence.…”

“…They provide a natural extension to the DS paradigm by incorporating multiple independent sensors with distinct observation models and noise characteristics. Prominent application domains for DDSs are wireless sensor networks [26], [27] and multiagent systems [28]. The mathematical foundations of DDSs provide a generic framework for modeling systems with multimodal sensory input.…”