Carrier Frequency Offset Estimation for OFDM Systems using Repetitive Patterns

This paper deals with Carrier Frequency Offset (CFO) estimation for OFDM systems using repetitive patterns in the training symbol. A theoretical comparison based on Cramer Rao Bounds (CRB) for two kinds of CFO estimation methods has been presented in this paper. Through the comparison, it is shown that the performance of CFO estimation can be improved by exploiting the repetition property and the exact training symbol rather than exploiting the repetition property only. The selection of Q (number of repetition patterns) is discussed for both situations as well. Moreover, for exploiting the repetition and the exact training symbol, a new numerical procedure for the Maximum-Likelihood (ML) estimation is designed in this paper to save computational complexity. Analysis and numerical result are also given, demonstrating the conclusions in this paper.