针对正交频分复用(OFDM)系统现有频偏估计时域方法中估计范围与估计精度的矛盾,提出一种基于时域训练序列的多步频偏估计新算法.该算法采用多步逐次逼近,在每一步估计中消除相位模糊,在不缩小估计范围的前提下不断提高估计精度;最后利用最佳线性无偏估计(BLUE)方法将各步估计值进行最佳合并,进一步提高估计精度.该算法用简单的修正取代搜索,理论分析和仿真表明在保证估计精度的同时,可显著降低大频偏估计的计算复杂度.该算法的主要运算是并行的,当用于高速通信时优势更明显.
For the existing time-domain methods of carrier-frequency offset estimation for orthogonal frequency division multiplexing(OFDM) systems,estimation range is enlarged at a cost of estimation accuracy reduction,or vice versa.A novel multistage frequency offset estimation algorithm based on time-domain training sequence was proposed to avoid the conflict between estimation accuracy and estimation range The algorithom gradually approached through multiple stages and removed the phase ambiguity in every stage,so improved the estimation accuracy continuously.Then the estimates obtained in all the steps were combined by carrying out the best linear unbiased estimation(BLUE) to further improve the estimation accuracy.Theoretical analysis and simulation show that by replacing heavy search by simple modification,the proposed algorithm can reduce significantly the computational complexity for estimation of large frequency offset while maintains good performance.The main operations of this algorithm can be carried out in parallel,thus the algorithm can be applied to high-data-rate communication systems.