An Improved Frequency Offset Estimation Algorithm for IEEE 802.11a System

This paper addresses the problem of the contradiction between estimation accuracy and estimation range of the frequency estimation in OFDM. According to the packet structure of IEEE Standard 802.11a and the characters of short training symbols, a new frequency offset estimation algorithm is presented. A two stages strategy based on auto-correlation algorithm is used to solve the contradiction between estimation accuracy and estimation range. The coarse frequency offset estimating bases on the identity of half cyclical short training symbol while the exact frequency offset estimating is done by the cancellation operation of long training symbols. The simulation based on 802.11a PHY model is done. And the simulation result shows that this algorithm not only overcome the influence of carrier frequency offset but also eliminate the carrier frequency offset. Comparing with the traditional algorithm, proposed algorithm can reduce the complication of the calculation while the estimation range can be expanded to 2 times. The results of the performance comparison are presented in terms of mean and mean-square error (MSE) obtained by simulations. The new algorithm is examined in Gaussian and Multipath fading environments for comparison. Furthermore, the precise estimation can be acquired as well.