Channel Tracking Based on Neural Network and Particle Filter in MIMO-OFDM System

Abstract

This paper proposes a novel channel tracking method based on radial basis function neural network (RBFNN) and particle filter (PF) in multiple input multiple output-orthogonal frequency division multiplexing (MIMO-OFDM) system. First, we use the RBFNN to obtain the initial values of the MIMO channels, and then apply the PF method to track the variation of the channels. The fading channels are modeled as autoregressive (AR) process and the transmit signals are encoded by space-time block code (STBC) scheme. Simulation results show that the proposed method has effective tracking performance.