基于ML和ESPRIT方法的OFDMA上行链路频偏估计算法

Carrier-Frequency Offset Estimation for OFDMA Uplink Based on ML and ESPRIT Method

由于载波频偏会给正交频分多址接入OFDMA系统带来子载波之间干扰,从而造成多用户之间的干扰,导致系统性能下降。该文提出一种基于ESPRIT和ML方法联合估计的频偏估计算法。该算法首先使用ESPRIT方法估计出多个可能的频偏构成的子集,然后使用最大似然估计方法在这个有限子集合中搜索出估计的频偏。该算法解决了使用似然估计进行多维搜索的问题,大大降低了算法的复杂度,同时解决了多个频偏估计的问题。

Carrier Frequency offsets in OFDMA system will introduce InterCarrier Interference （ICI） and consequently degrade the system performance. A new frequency offset estimation algorithm for OFDMA system based on the combination of ESPRIT （Estimation of Signal Parameters by Rotational Invariance Techniques） and the Maximum Likelihood （ML） estimation method is proposed. The subset of the frequency offsets first is estimated by ESPRIT, then the frequency offset of each user is identified in the subset by the ML estimation method. The algorithm greatly decreases the complexity of N-dimensional search processes of ML method and efficiently solves the problem of multi-frequency offsets estimation. Key words：