Iterative joint data detection and channel estimation employing LMS algorithm in block flat fading environments for multi-stream system

Abstract

BLAST (Bell Labs layered space-time) provides high capacity wireless communications in rich scattering environments. We investigate the problem of channel estimation of VBLAST (vertical BLAST) over a block flat fading environment. We proposed an effective iterative joint data detection and channel estimation method. At the beginning of a block, we apply a ML (maximum likelihood) estimator to achieve the initial channel estimation, and then at every time interval, the new channel estimation is obtained by performing the LMS (least mean square) algorithm. At the end of the block, a low-pass filter is used to effectively reduce the effects of noise to obtain a more accurate channel estimation value, which is the initial channel estimation for the next iteration. By computer simulation, the performance of the proposed channel estimator proved to be much better than that of the conventional channel estimator.