The ML algorithm of frequency offset estimation for OFDM systems based on pilot symbols and decision data

In order to track the real-time variation of the multipath channels, pilot symbols are often inserted into time-frequency grid to estimate the channel response in OFDM systems. These pilot symbols can be used to estimate the frequency offset, and the main merit of this method is that good frequency offset estimation can be obtained without additional data. Further analysis for this method is given, the approximate ML estimation result in time domain is derived, and furthermore, likelihood function of different style in frequency domain is given, and analytical estimation result is also derived. Compared with the original estimation algorithm, new algorithm can reduce the computational complexity greatly, and improve the accuracy of the estimate greatly. The algorithm can work effectively even if the signal to noise ratio is very low.