Performance Comparison of Two LMS Adaptive Beamforming Algorithms Based on Pre-FFT Combining for OFDM

摘要

In this paper, we compare the performance of two least mean square (LMS) adaptive beamforming algorithms based on pre-FFT combining for Orthogonal Frequency Division Multiplexing (OFDM). One algorithm is with the frequency-to-time pilot transform while the other one does not have such transform. The performance investigations verify that the algorithm with the transform is capable of giving more accurate estimation of the antenna weights. However, this superior characteristic can not contribute to the improvement in either the beam pattern generation or the BER performance. The second observed result is that for LMS adaptive beamforming based on pre-FFT combining, the increase of the pilot symbol number can not achieve the performance improvement. The investigation results give some good suggestions for the engineering implementations.