Low Snr Environment, The Distributed Mimo System Frequency Synchronization Technology

Multiple-Input Multiple-Output (MIMO) which is widely used in Long Term Evolution (LTE) has become attractive due to its advantages of interest in filed of wireless communications. It can be found that the prominent direction about MIMO technology is the distributed MIMO system from the trend of research. It can be set up transmitting and receiving antennas according to the specific needs for its higher capacity.As the transmitting and receiving antennas may be located in different geographic locations, the signals transmit through different channels, so the distributed MIMO system puts up higher requirements for time and frequency synchronization. This problem must be solved for its advantages. But how to overcome the multiple antennas interference (MAI) to achieve effective synchronization performance has become the focus of research, especially in low Signal to Noise Ratio (SNR) environment. In this paper, it has done some research and put forward some new frequency synchronization algorithms under the distributed MIMO system. These algorithms suit the frequency synchronization especially for low SNR environment. At first, Chapter 1 introduces the research background and the status of MIMO technology in the field of wireless communications and illustrates the necessity to research. And then, it mentions some basic principles and strengths briefly of MIMO system, and identify the main issues and the basic objective of the paper. Next, it describes the main contribution and makes arrangements for the content of this paper.At second, Chapter 2 summarizes the main existing frequency synchronization algorithms. It states that the frequency synchronization of the distributed MIMO system is actually a multi-frequency-offset estimation. The existing algorithms can be divided into the Maximum Likelihood (ML) and quasi ML principle. However, the difference is that there exists MAI in the distributed MIMO system and how to conquer the MAI is a fundamental problem to be solved. Furthermore, the Expectation-Maximization (EM) kind of solving ML estimation algorithms which can overcome the MAI have been influenced by initialization and local convergence, especially under the SNR environment. So how to implement the frequency synchronization effectively is another problem for this situation.For solving the problems, the Chapter 3 focuses on the optimization frequency synchronization algorithms which are EM kind algorithms and can break through MAI and achieve the Cramer–Rao Low Bound (CRLB). Under low SNR environment, those algorithms are difficult to converge to the global optimum value due to the sensitivity of initialization and local convergence. For this reason, this paper addresses a sphere-decoding algorithm based frequency-offset estimation for the distributed MIMO System for finding the closest lattice point to the received signal within a sphere of radius limited by the asymptotic CRLB. Because it is independent with grads of target function about frequency offset, simulation results show that the proposed algorithm can achieve the CRB and improve the drawbacks of existing algorithms.Actually, frequency offsets to be estimated are held unfixed and changed by the environment factors such as temperature, humidity and so on. For this reason, the chapter 4 proposes a maximum a posteriori (MAP) estimation algorithm for distributed MIMO system based on prior information of frequency offsets. Under the generalized maximum entropy principle, it is widely recognized that the prior distribution to be choose should make the entropy maximal when there is no definite prior information. Furthermore, it derives a sub-optimal MAP estimation algorithm called quasi maximum a posteriori (QMAP) algorithm. It uses a revised approach to reduce the complexity due to priori information. Compared with the ML estimation, simulation results illustrate the performance of MAP estimator achieve the CRB and the proposed algorithm can improve the efficiency of synchronization obviously, especially in low SNR environment.This paper makes an in-depth research on frequency synchronization which is one of the key technologies to implement the distributed MIMO system. As will be shown, the proposed algorithms have great potential for theory and application especially under low SNR environment.