OFDM水声通信中非一致多普勒补偿技术研究

随着海洋开发的不断深入，对水下移动通信的有效性和可靠性的要求也越来越高。由于水声信道的严重带宽限制，在无线通信领域已成功应用的OFDM技术以其带宽利用率高的特点成为水声通信领域内的热门研究课题。但是水声信道时变、空变和通信双方相对运动产生的多普勒会破坏OFDM正交性从而影响系统性能。同时由于水声OFDM中多普勒频偏相对于子载波间隔要比无线通信中大上千倍，因而目前还有行之有效的方法来补偿OFDM水声通信中的非一致多普勒问题。本课题研究并提出基于自适应迭代的宽带多普勒补偿方法，解决水声OFDM中的非一致多普勒初步补偿问题，该方法在迭代过程中抑制自身与相邻载波的干扰，达到初步补偿的目的。提出基于压缩感知的宽带多普勒补偿方法，解决残余多普勒的补偿问题，该方法能够对信道每条路径的幅度、时延和多普勒进行估计，性能优于利用训练序列估计信道的方法。开展湖上试验和海上试验，验证并改进算法性能。

With ocean development increasingly thorough, requirements of validity and reliability for underwater mobile communication are also getting higher and higher. Due to the severe underwater acoustic channel bandwidth limitation, the OFDM technology with its high bandwidth utilization characteristics, which has been successfully applied in the field of wireless communications, becomes a popular research topic in the field of underwater acoustic communication. The performance of the underwater acoustic communication is largely dependent on the underwater acoustic channel propagation conditions because the underwater channel is time-varying, space-varying, multipath fading channel whose bandwidth is severely limited. Also, in underwater acoustic OFDM system the Doppler frequency shift relative to the subcarrier spacing is a thousand times higher than that in the wireless communication field. Thus, so far there aren't any effective methods to handle the problem of compensating non uniform Doppler of the OFDM underwater acoustic communication. In this research, we propose a kind of broadband Doppler compensation method based on adaptive iteration to solve the non-uniform Doppler preliminary compensation in the underwater acoustic OFDM system. This method can inhibit interference among adjacent carriers in the iterative process, and finally achieve the purpose of the preliminary compensation.To deal with the residual Doppler frequency shift, a broadband Doppler compensation method based on compressed sensing is proposed. The method can estimate the amplitude, time delay and Doppler frequency shift of each path, which is superior to the channel estimation approach based on the training sequences. At last, carry out experiments in lakes and seas to validate and improve the performance of the algorithm.