Blind source separation of chaotic signals in wireless sensor networks

Abstract

Chaotic signal is essentially a nonlinear and non-Gaussian signal, which involves signal quantization when used in wireless sensor networks (WSNs). It makes the blind source separation of chaotic signal in WSNs more difficult to address. To solve the problem, we propose a new source separation algorithm based on cubature Kalman particle filter (CPF) in this paper. First the probability density function of the observed signal is derived and the optimal quantization is used; this can achieve the optimal quantization of signal under the limited budget of quantization bits. After that, the algorithm uses cubature Kalman filter (CKF) to generate the important proposal distribution of the particle filter (PF), integrating the latest observation and improving the approximation to the system posterior distribution, which will improve the performance of the signal separation. Simulation results show that the algorithm can separate mixed chaotic signal effectively, it is superior over the unscented Kalman particle filter (UPF) counterpart in accuracy and computation overhead. The running time is 88.77% compared to the UPF counterpart.