Performance analysis of least mean square sample matrix inversion algorithm for smart antenna system

Abstract

Beamforming is a signal processing technique that has the ability to direct the main beam toward signal-of-interest and the null toward signal-of-not-interest without moving antenna array. This technique is achieved by using algorithms which change the amplitude and phase of array pattern continuously. This paper presents hybrid algorithm that is a combination of two algorithms, least mean square algorithm and sample matrix inversion algorithm. The hybrid algorithm (least mean square/sample matrix inversion) is applied on an array of dipoles to overcome the shortcomings of existing algorithms for a robust smart antenna system. To focus the main beam toward the desired direction and place the null in the direction of the interference signals, the weights of the inversion matrix in sample matrix inversion are calculated and these weights will be the initial weights in least mean square algorithm. The merit of this approach solves the convergence speed problem of the least mean square algorithm as well as the computation intensive exists in sample matrix inversion algorithm and decreases the least mean square error. The simulation results indicate the analyses of least mean square, sample matrix inversion and hybrid algorithm performances .These techniques are compared and verified using MATLAB.