The Study Of Adaptive Algorithms In Smart Antenna

In recent years, with the rapid development of mobile communications, the number of users is of great increase, which leads to the increasingly stress of spectrum resource. Meanwhile MAI (Multiple Access Interference), CCI (Common Channel Interference) and multipath fading limit the further improvement to the channel capacity and quality of communication. Since the smart antenna technique could overcome the preceding problems with its space characteristics, the use of smart antenna could not only guarantee the quality of communication, but also raise the region of covering and the system capacity.The term "smart" means the antenna can be adaptively modified the weighted values according to the needs of the users and the propagation environment in certain range. The adaptive modification is controlled by adaptive algorithms, so that the adaptive algorithm is the core of smart antenna technology. The performance of the whole system is directly affected by the convergence performance and computation complexity of the adaptive algorithms. The algorithms that have good convergence performance typically require large computation and complex structure.Therefore, the low computation and concise implementation of adaptive algorithms are always pursued by researchers.In this paper, the principle and work version of smart antenna are proposed firstly focusing on the adaptive work version, and then the features, mathematical model and simulation methods of mobile communication are presented. Two classic adaptive algorithms (LMS and RLS) are discussed as well as two modified adaptive algorithms (HRLS ad NLMS). The convergence performances of these algorithms are analyzed and verified by simulation results. Moreover, two novel proposed adaptive algorithms are introduced. One is the modification for HRLS, and anther is a modified RLS for real-value modulation system. Simulation results verity that the proposed HRLS has stronger ability on interference suppression than RLS. Furthuremore, it keeps the advantage of lower complex computations and less total convergence time. And a modified RLS for real-value modulation system reduces the convergence time by half than RLS, simulation results show that the space resolution of the proposed RLS is higher than that of RLS.Finally, we realize the flat fading channel and the frequency selective fading chanennl by simulation, and inspect the performances of the preceding algorithms. Simulation results show that the performences of the algorithms become worse when the doppper shift or the number of the interferences and multipath increases. Thereinto, the proposed RLS for real-value modulation system has the strongest ability of the interferences and multipath suppression, and the ability of the proposed HRLS algorithm is stronger than that of HRLS algorithm which has the wost ability.