A simple LS algorithm for improving ESPRIT-based blind CFO estimations in OFDM systems

The estimation of carrier frequency offset (CFO) is an important issue for OFDM systems. Many CFO estimation methods have been proposed in the past. In particular, two ESPRIT-based methods were introduced for blind CFO estimations. These ESPRIT-based methods can provide satisfactory performance at a reasonable implementation cost. In this paper, we propose a least-squares (LS) algorithm for improving the performance of these ESPRIT-based methods. Simulation results show that the proposed LS algorithm, although very simple, can significantly improve the accuracy of CFO estimation.