Capacity improvement techniques for cellular communication systems such as smart antenna systems are becoming more important as demand on mobile communication services increases. This paper analyzes effects of different wavelet funtions applied to the received signal vector by using the stationary wavelet transform (SWT) before the direction of arrival (DOA) estimation. The Multiple Signal Classification (MUSIC) algorithm is used for the DOA estimation. Resolution of the MUSIC algorithm for uniform linear arrays decreases as impinging signals get close to the horizon, due to array steering vector. Enhancement of the SWT before the DOA estimation is to enhance this poor resolution problem. Different wavelet functions effect the performance of this method so this paper statistically analyzes these effects by simulating the received signal vector for signals arriving from close horizon angles and adding a random level of noise to this vector. After the modelling of the received vector the SWT is applied by using different wavelet functions and results are compared after 100 simulations.