OFDM系统中对最大似然算法频偏估计的改进

深入研究了基于循环前缀(Cyclic-Prefix,CP)的最大似然算法(ML)算法。针对这个问题,对基于CP的ML算法做了如下改进,改进的算法同时对连续3个符号进行ML估计。符号定时同步时,当前符号同步信息能够通过前面2个符号的同步信息得到校正。在进行频偏估计时,通过使用观察窗,在窗口内进行连续频偏值的均值运算,对频偏估计值误差进行矫正。

Improvement of Frequency Offset Estimation of ML Algorithm in OFDM

The Maximum Likelihood（ML）method based on Cyclic Prefix（CP）is studied thoroughly.Frequency offset estimation is not accurate due to the wrong estimation of symbol timing.To solve this problem,the paper proposes the improvement of ML algorithm based on CP.The improved algorithm estimates three consecutive symbols at the same time.When considering symbol timing synchronization,the timing estimation of the current symbol could be corrected through the first and second symbols.The algorithm uses observation window to reduce the frequency offset error,in which the average of consecutive frequency offset estimation values of three symbols is obtained during frequency offset estimation.