Joint evolutionary spectrum and autoregressive-based approach to modeling non-stationary flat fading channels

摘要

Modeling of wireless channels, especially non-stationary fading channels, is important for design and performance analysis of wireless communication systems. Recently, we proposed a new approach to modeling non-stationary fading channels, based on the theory of evolutionary spectrum (ES). In this paper, we develop a time-varying autoregressive (AR) model for a non-stationary flat fading channel; specifically, we develop a method to determine the time-varying coefficients of the AR channel model, given the ES of a non-stationary process. Furthermore, with the ES theory, we develop a trace-driven time-varying AR channel simulator to generate a non-stationary flat fading process. Simulation results show that the ES of the channel gain process produced by our joint ES-and-AR-based channel model agrees well with the user-specified ES, indicating the accuracy of our joint ES-and-AR-based channel model.