Wavelet BEM based channel estimation over rapidly time-varying channels

Abstract

This paper proposes a novel basis expansion model (BEM) to conduct channel estimation for single carrier single input single output (SC-SISO) systems over rapidly time-varying channels. Compared to complex exponential (CE) and polynomial BEM, wavelet BEM is less vulnerable to Doppler shift. Simulations show that the proposed wavelet BEM outperforms CE and polynomial BEM with about 2.5dB and 5.0dB respectively in terms of the mean square error-signal noise ratio (MSE-SNR) performance.