A Novel Channel Estimation Method for TD-SCDMA System

The B. Steiner estimator used in the TD-SCDMA system assumes that the channel response during one timeslot is constant, which results in big errors under the fast varying channel condition compared with the real channel response. In order to solve this problem, a novel channel estimation method is proposed for the TD-SCDMA system. Every timeslot is separated into several segments, the multi-segment estimation and multi-slot LSM fit is used to improve the accuracy of channel estimation. By comparing this two methods and analyzing their effects to detection, especially for the joint detection (JD) algorithm of TD-SCDMA system, simulation results show that the novel channel estimation is more close to the real channel response, which can largely improve the system performance, especially for the fast varying channel environment