Abstract:
The residual frequency offset (RFO) may lead to constellations rotation, and this rotation is related to the time. The longer the time is, the larger the rotation will be. If the RFO cannot be corrected for long time, the decision error may occur because of constellations rotating to other quadrant, which will severely degrade the performance of OFDM systems. In order to correct the RFO, this paper proposes a RFO tracking algorithm based on constellation characteristics for OFDM systems. Based on the assumption that the channel frequency response has been estimated and keeps constant within one frame, the proposed algorithm is accomplished by using the properties of constellations rotation. At first, constellations of every OFDM symbol are divided into four subsets, and then phase error is extracted by solving phase angle of every subset’s average. Finally, the frequency offset compensation in the domain is achieved by using a two-tap loop filter. The input SNR of the loop may be improved effectively, because the CFO estimation is performed by using all data subcarriers and averaging every subset. With proper loop parameters, a CFO less than 5% of the subcarrier spacing may be well tracked for OFDM systems with QPSK modulation. While for OFDM systems using QAM modulation, the CFO tracking range may be reduced to ±0.01 of subcarrier spacing. The simulation results show that the tracking performance of the proposed algorithm is better than the one of ML estimator based on cyclic prefix, and is very close to that of estimator based on pilot tones.