LMS beamforming for OFDM system in wavelet domain

Frequency selective fading effects in OFDM based high speed wireless communication systems can be highly mitigated by using adaptive antenna arrays. In addition, the use of discrete wavelet transform (DWT) processing in OFDM systems can successfully counter the degrading effect of inter symbol interference (ISI) without the need of cyclic prefix (CP), and hence conserves band width and improves the spectral efficiency. In this paper, an OFDM system utilizing DWT based least mean square (LMS) adaptive beamformer is considered. The use of the wavelet transform reduces the signal space dimension and hence results in low complexity in addition to data denoising. Time saving in signal beamforming processing can be achieved by using different levels of decomposition in the wavelet transform, with acceptable accuracy. Theoretical analysis and computer simulations demonstrate how the suggested scheme, with different wavelet families and different decomposition levels, achieves better beamforming performance with faster convergence, as compared to conventional adaptive beamforming.