Research On Joint Detection For TD-SCDMA Downlink

The capacity and performance of a code division multiple access system is essentially dominated by multiple access interference and inter symbol interference. Therefore, the suppression and cancellation of MAI and ISI is a critical issue to improve performance and increase capacity in the CDMA system. Joint detection is applied to solve the problem in TD-SCDMA and its core thought is that sent to all users of the data character as a single-user sending data sequence, in the receiving detection, known users spreading codes and the channel impulse response (CIR) valuation, MAI will be processed as ISI.But these already existing algorithms calculating complicated, greater part of joint detection algorithms only can be applied for the bastion station. More effective algorithm is deficient in user equipment. Joint detection can execute the others procedure after it must know all spreading code information in one time spacing. While in downlink, user equipment couldn't know the others users'information of spreading code. ZF-BLE-SD and MMSE-BLE-SD successfully solved these problems. Among the two single user algorithms, MMSE-BLE-SD performance is better than ZF-BLE-SD.Discusses in detail the principle of joint detection algorithm of TD-SCDMA communication system, especially the MMSE-BLE-SD (single-user minimum mean square error detection block equilibrium) algorithm Principle. At the same time , described the principle of independent component analysis and its application in multi-user detection of CDMA communications system.Then inferenced Independent component analysis in the TD-SCDMA communication systems used is reasonable.Last the paper present using ICA as an additional processing element attached to the receiver of the downlink of TD-SCDMA system after referring to ICA methods in CDMA MUD applications. MMSE- BLE-SD is applied in downlink and the same time according to its characteristic modified the model of downlink simulation and complex FastICA algorithm applied in ICA and test and verify the algorithm throught computer simulation.The simulation results show that, ICA as the post-processing modules of Joint detection, in the circumstances of high signal to noise ratio, the performance of the system was improved.