Rician channel characterization in OFDM systems

Rician model is a stochastic channel model which is used for multipath propagation when line of sight (LOS) signal is much stronger than the other signals received from the other paths. In this paper we focus on Rician model of fading channel in Orthogonal Frequency Division Multiplexing (OFDM) systems and evaluate the effect of LOS signal on channel characterization. We show that LOS signal makes correlation between consequent covariance matrices which are used to estimate channel response in OFDM systems. As a result, a weighted average between covariance matrices can improve statistical properties and increase the accuracy of estimation in limited number of OFDM symbols. Motivated by the above idea, we propose a simple updating technique to get optimized performance on channel characterization. We investigate the asymptotic performance of the proposed estimator, i.e. normalized mean square error (NMSE) and Cramer-Rao bound (CRB), for large numbers of independent observations. The results clearly show the effectiveness of our proposed algorithm in estimating channel response as well as its improvement over the conventional techniques.