Doppler Frequency Offset Estimation Algorithm For TD-SCDMA System In High-speed Environment

TD-SCDMA with independent intellectual property rights of third-generation mobile communication systems, smart antennas, joint detection and relay switching technology compared to other 3G standards it has a higher spectrum efficiency and greater system capacity . TDD technology, the user channel environment detection and adjustment of non-consecutive in time, various algorithms of users supported by the highest mobility rate of only 120 km / hour. In high-speed environment, there are frequent cell switching, switching in the residential area adjacent to the existence of a strong cell interference users. High-speed environment, the channel estimation algorithm must take into account the impact of residential neighborhood, the traditional channel estimation algorithm for single-cell environment at high speed on the user channel is not an accurate estimate of the environment; terminals and base stations, relative movement between the Doppler frequency generated shift directly affects the phase of the received signal, the middle time slot by the user Doppler shift estimation midamble codes have been the traditional algorithm can not achieve the correct signal demodulation.Based on the above problem, we studied the high-speed TD-SCDMA system in the context of channel estimation and compensation algorithm estimates the Doppler shift, the main work:1. Proposed a multi-cell joint channel estimation algorithm for high-speed environment.The algorithm is based on the size of the power users choose a powerful neighbor cell interference, adjacent to the residential user and the user community with strong interference into the channel estimation matrix channel estimation. Simulation results show that multi-cell joint channel estimation algorithm than the steiner algorithm has higher accuracy.2. Proposed a signal model based on joint detection, the use of data fields for the Doppler shift estimation algorithm. The method for the initial training sequence in the use of frequency estimation, the use of multi-cell joint channel estimation and channel response obtained by the Doppler shift estimate of the initial training sequence at both ends of the data domain demodulation; Finally got demodulation user data a second frequency offset estimation. Simulation comparison, in the same SNR conditions, the algorithm than the traditional use of training sequences to estimate the compensation algorithm has a lower error rate.