基于ML和导频联合的OFDM同步研究

Research of Synchronization of OFDM based on ML and Pilots

研究了如何将最大似然算法(ML)应用于OFDM系统,并且对该算法进行了改进。ML算法是利用相邻符号中循环前缀的相关性进行时频同步估计,但是其在进行频偏估计时只能进行小数倍频偏估计。当采用最大似然算法联合插入导频做系统同步时,就能同时进行小数和整数倍频偏估计。采用联合算法后,系统的误码率有了明显的下降,接近于理想同步的效果。仿真结果验证了该联合算法的有效性和实用性。

In this paper, we do the research how to use the maximum likelihood （ML） algorithm to resolve on the synchronization of OFDM system. Besides, we improve the algorithm. The algorithm exploits the correlation of cyclic prefixes in adjacent symbols to perform time and frequency synchronization,but it only acquires fractional frequency offsets. We can acquire fractional and integral frequency offsets when joint ML algorithm and pilots. After adopting the joint algorithm,BER of the system will acquire evident decreases ,close to ideal synchronization. Simulation results verify its validity and practicability.