A statistical approach to the analysis of DS/CDMA cellular systems employing RAKE receivers and sectorized antennas

Abstract

RAKE receivers and sectorized antennas are used in direct-sequence/code-division multiple-access (DS/CDMA) cellular systems to improve the system performance. This paper presents a statistical method for analyzing the performance of DS/CDMA cellular radio systems employing RAKE receivers and sectorized antennas. Average bit error rates in the system are estimated considering the multipath fading effects of the environment. (The fast fading is assumed to be Rayleigh distributed, and the distance-dependent means of the multipath components have an exponential power delay profile.) The analysis of RAKE receivers quantifies the performance improvement that could be achieved by increasing the number of RAKE fingers. Sectorized antennas improve the system performance by reducing the interference at the receiver. In a perfectly sectorized system, assuming three sectors per cell, the capacity of the system can be improved by a factor of three. However, due to the imperfection in practical antennas, it is not possible to achieve this improvement. The performance of systems employing practical sectorized antennas (with finite front-to-back ratios and overlapping sectors) is compared with the performance of perfectly sectorized systems. The analysis shows that the incremental performance improvement diminishes with each incremental increase in the number of RAKE fingers. Performance degradation due to finite front-to-back ratio is shown to be insignificant for practical values of the front-to-back ratio of sectorized antennas. However, the reliability of mobile reception can be degraded significantly in areas where adjacent sectors overlap.