OFDM水声通信多普勒频移的最大似然估计

Maximum likelihood estimation for Doppler shift of OFDM underwater acoustic communication

针对水声信道的时间、空间及频率变化特性,以及由于发射机和接收机之间相对运动产生的多普勒频移对OFDM子载波间正交性的破坏问题,提出一种OFDM水声通信系统多普勒频移的估计方法,用于估计和补偿接收机的多普勒频移.该方法以OFDM数据块为研究对象,假设在一个OFDM数据块内多普勒因子不随路径变化,采用最大似然估计算法估计多普勒频移,通过仿真分析系统多普勒因子临界值,并将多普勒因子估计方差与克拉美罗界进行比较.结果表明：系统存在一个信噪比范围并-15～0 dB,此时估计方差最接近于克拉美罗界.该方法还可以给出系统多普勒因子的临界值,能够保证在不同信道条件下估计参数的一致性,提高了多普勒因子的估计精度.

This paper proposes an improved method capable of estimating and compensating the Doppler effects at the receiving end,an approach designed for working with the metabolic time,space and frequency of underwater acoustic （UWA） channel,and above all overcoming the notorious effect on the orthogonality among subcarriers by the Doppler frequency shift resulting from the relative motion between transmitters and receivers.The proposed method building on the OFDM data block and on the assumption that Doppler factor does not change with the path during one block works by estimating the Doppler frequency shift using the maximum likelihood estimation algorithm;providing the simulation analysis of the critical value of Doppler factor,combined with the maximum likelihood algorithm;and ultimately comparing the variance of the factor with the Cramer Rao bound（CRB）.The results show that the system features a rang of SNR（-15-0 dB provided in this paper）,allowing the estimation variance to be the most close to the CRB.The method capable of giving the threshold value of Doppler shift,and thereby ensuring the consistency of the estimated parameters in different channel conditions could contribute to an improvement in the estimation precision of Doppler factor.