3 Gpp Lte Downlink Link The Study Of The Initial Synchronization

Downlink initial synchronization is the first fundamental procedure in the cell search. In this thesis we investigate the downlink initial synchronization for the3rd generation partnership project (3GPP) long term evolution (LTE) system. We review the progress of3GPP LTE standardization, describe the LTE physical layer procedure generally, and introduce the synchronization channel (SCH) and SCH sequences in detail.We first investigate the carrier frequency offset (CFO) estimation for3GPP LTE downlink initial synchronization. Through analysis and computer simulation, three CFO estimation schemes including the primary synchronization channel (P-SCH) correlation based scheme, cyclic prefix (CP) based scheme and multiple signal classification (MUSIC) based scheme, are reviewed in terms of normalized mean square error (MSE), respectively. Furthermore, we propose two efficient CFO estimation schemes:one is the extended P-SCH correlation based scheme, the other combines the traditional P-SCH correlation based scheme and CP based scheme adaptively according to both different signal-to-noise ratio (SNR) and CFO regions. Simulation results show that, both of the proposed schemes can achieve better normalized MSE performance over additional white Gaussian noise (AWGN) channels and frequency selective fading channels.We also study the detection algorithm of the secondary synchronization channel (S-SCH) for3GPP LTE downlink system. Considering the optimal maximum likelihood sequence detection (MLSD) algorithm as the guideline, a low complexity detection algorithm is proposed by utilizing the special structure of the S-SCH sequence to reduce the number of candidate S-SCH sequences. Fast Hadamard transformation is applied to further reduce the computational complexity when calculating the likelihood function values. Analysis in our work illustrates that the proposed algorithm reduces the computational complexity significantly when compared to the MLSD algorithm. Simulation results show that our proposed algorithm also achieves a good tradeoff between correct detection probability and the computational complexity.