Adaptive semi-blind channel estimation for OFDM systems

Orthogonal frequency division multiplexing (OFDM) with high throughput is an attractive technique for broadband digital broadcast system (DBS) and high rate wireless local area network (WLAN). In this paper, an adaptive semi-blind channel estimation method based on the second-order statistics of the received signal is proposed. Employing this method, the cyclic prefix (CP) of OFDM symbol can be removed, which results in improved transmission throughput. This approach also ensures a fast convergence and moderate complexity by taking advantage of training symbols which are available in most OFDM systems. Compared to blind channel estimation methods, the proposed semi-blind method needs lower computation. Monte Carlo simulations show that it also outperforms these blind algorithms in convergence rate.