Research On Joint Detection Algorithm And Its Implementation In TD-SCDMA System

Communications in China has been developed for more than one hundred years. Accepted by International Telecom Union as one of the three major standards for mobile communications technology of the third generation with WCDMA and CDMA2000, TD-SCDMA is our first international communications standard with independent intellectual property rights. TD-SCDMA adopts working manner of TDD requiring no matching frequency, and multiple access combing with FDMA, TDMA and CDMA. Besides, it makes use of smart antenna, joint detection, synchronous CDMA, low chip rate, relay switching, adaptive power control, software radio and many other advanced technologies. In aspects of spectrum flexibility, spectrum utilization rate, environment adaptation and equipment cost, it has some distinct advantages over other 3G systems. The paper puts emphasis on the analysis of joint detection. By treating all the interfered clients as the useful signal and making some process, using the recognized information from client signals to eliminate or reduce the multi access interference, multi-path interference and near-far effects, the system capacity is improved.At the very beginning, the paper introduces the basic format and modulation manner of the physical channel in TD-CDMA, and then discrete transmission model and matrices of the system are established combining with features of TD-CDMA. Signal models are constructed respectively according to the receiver and accepter. The structure of the system matrices is introduced with particular concern. Basic principle of joint detection and those joint detection algorithms are also described, compared and analyzed in the paper. Finally, by using MATLAB software simulation platform, the paper puts emphasis on simulation analysis and comparison of ZF-BLE and MMSE-BLE algorithms in Reyleigh fading channel model. Meanwhile, the performance differences of the two algorithms with different parameters are tested. The influence of the increasing clients on the deterioration of the algorithm performance is also proved.