Performance analysis of kernel based adaptive beamforming for smart antenna systems

Abstract

In this paper, kernel based adaptive algorithms for smart antenna systems are considered to achieve the best weight vector for beamforming by iteration process. Smart antenna systems radiate not only narrow beams towards desired signal directions, but also place deep nulls in the undesired signal directions, thus optimizing the signal quality, and enhancing the coverage range. The performance of the algorithms are studied with parameters, like convergence speed and minimum mean square error. Based on simulation results, it is proved that kernel based adaptive algorithm provides better performance in terms of weight convergence, null steering and power gain over that of the traditional least square algorithm. Therefore, the kernel based algorithm gives a cost effective solution in mobile communication under dynamic channel environment to enhance capacity and range.