Multisource Prediction Under Nonlinear Dynamics in WSNs Using a Robust Fuzzy Approach

This paper focuses on addressing the robust source prediction problem in wireless sensor networks (WSNs). In particular, the nonlinear interaction of multiple sources is handled by the fuzzy interpolation technique, which has not been fully investigated in networked data fusion systems yet. In this study, no knowledge of the joint distribution of sensor observations is assumed. A robust H_∞ design using fuzzy interpolation from a game theory viewpoint is proposed. The fuzzy approximation error for the nonlinear dynamics of multiple sources is incorporated into the design. In evaluation, a theoretical analysis on the performance of the proposed multisource prediction method is provided.