Research And Implementation Of Anti Doppler And Channel Estimation Techniques For Underwater Acoustic Speech Communication

As an indispensable technology in the exploration and exploitation of marine resources,underwater acoustic communication has gradually become the hot spot of research. As one of the most direct and effective communication, speech communication plays an important role in underwater acoustic communication. Due to the limited available bandwidth of underwater acoustic channel, and the existence of serious noise, multipath time delay and Doppler shift,the research and implementation of underwater acoustic communication has been very difficult. Orthogonal frequency division multiplexing (OFDM) technology is widely used in the field of underwater acoustic communication, with the high bandwidth utilization and strong ability to resist multipath. But OFDM is very sensitive to carrier frequency offset, and the Doppler frequency shift destroys the orthogonality between the OFDM sub-carriers,which seriously affects the performance and quality of communication. Therefore, it is very important to research and implement an OFDM underwater acoustic speech communication which can eliminate the influence of Doppler shift.This thesis mainly researches the key technologies of high quality underwater acoustic speech communication, which is based on the Doppler frequency shift. And the hardware implementation is carried out. Mainly completed the following work:1. The key technologies of underwater acoustic speech communication are researched.First, the 0.6kbps mixed excitation linear prediction (MELP) speech coding technology is studied, and through simulation, it is verified that the synthetic speech of the algorithm has high clarity and intelligibility. Then, the OFDM technologies are studied, and through simulating the convolutional coding of channel coding, the thesis comparative studies the performance of the convolutional encoding in different rates.2. Doppler estimation and compensation technologies of OFDM are researched. Firstly,the influence of Doppler frequency shift on the OFDM system is analyzed, and the principle of Doppler estimation and compensation is described. Then, by comparing the advantages and disadvantages of different Doppler estimation methods, this thesis proposes a new Doppler estimation algorithm which based on copy correlation and null subcarriers, and the Doppler compensation method is studied. Finally, through the simulation and pool test, the superiority of Doppler estimation algorithm which based on copy correlation and null subcarriers is verified.3. Channel estimation and equalization technologies of OFDM are researched. First, the system model of OFDM is analyzed, and the traditional LS channel estimation is studied.Then, using the sparse characteristics of underwater acoustic channel, the orthogonal matching pursuit (OMP) algorithm of compressed sensing is studied for channel estimation.By improving the algorithm, a new OMP channel estimation algorithm based on the noise variance is researched. Finally, the advantages and disadvantages of the channel estimation algorithms are verified by simulation comparison.4. The duplex and real time underwater acoustic speech communication system is implemented in the DM642 processing board. The performance of the system is verified by the electrical test and the pool test. The results show that the system can perform high quality speech communication, and the recovery of speech has a high clarity and intelligibility.