Abstract: For multi-path interference characteristics of wireless channel,the data-aided frequency offset estimation al-gorithm is improved based on the least squares(LS) error criterion.And a modified algorithm is proposed according to the correlation of the training sequence or the preamble symbols.The simulation results under specifications of ZigBee show that the mean square error(MSE) of the modified algorithm is better than that of the BAF(best angular fit) algorithm.When SNR is greater than 9 dB,the accuracy of frequency offset estimation of the algorithm comes up to the accuracy of the com-bination algorithm of coarse frequency estimation and fine frequency estimation.At the same time,the MSE of the proposed algorithm is close to Cramer-Rao low boundary(CRLB).Therefore,the algorithm can meet communication requirements for automatic meter reading system based on Zigbee technology in the metropolitan wireless channel environment.