The Iterative Synchronization And Equalization In Underwater Acoustic Communication

Different from radio communication and fiber-optical communication, underwateracoustic communication is limited by long multipath-delay, narrow available bandwidth,temporal and frequency variation, large Doppler, etc. Until recently, the design of underwatercommunication systems has mostly relied on the use of non-coherent modulation techniquessuch as FSK modulation. However, to achieve high data rates on the severely band limitedunderwater acoustic channels, bandwidth-efficient modulation must be considered. Thetypical problems that phase-coherent communication systems face are intersymbolinterferences and the sensitiveness to the variation of frequency and phase. This paper usesthe single carrier communication as the modulation way, studying the detecting algorithm toimprove the performance of detection, studying the channel estimation, passive time-reversalmirror, decision feedback equalizer, Turbo equalizer to solve the intersymbol interferences,using the phase-locked loop and the timing recovery to reduce the sensitiveness to thevariation of frequency and phase. The evaluation criteria for them are detection probability,system’s wrong bit rates, the output SNR, the tacking frequency or phase and the MMSEcurve.To test and verify the algorithms mentioned above, this paper use the Bellhop software toproduce the South Sea’s January channel (including shallow and deep), the estimated tank’schannel, G. Proakis’s channel B to simulate the receiver’s performance. Also, the tank’s datawhich is modulated by16QAM is processed and analyzed. The simulation results show thatthe passive time reversal based on channel-estimation can relieve intersymbol interferences,and the combination between time reversal and equalizer can reduce the system’s complexity,but the performance is affected by the modulated orders; The phase-locked loop can track andcompensate the frequency and phase variation, and timing recovery can solve the Dopplerproblem which makes the receiving signal compression and expansion, but the use ofphase-locked loop and timing recovery ask for higher SNR to preserve the system’s steady;Compared to the decision feedback equalizer, the turbo equalizer can work on lower SNR andreduce the wrong bit rates by iterative ways, but the iterative ways will introduce highercomplexity. The results of processing experimental data show that, in certain SNR, both thedecision feedback equalizer and the turbo equalizer can solve the intersymbol interferences insteady channel, the passive time reversal can relieve the intersymbol interferences from themultipath channel, but the advantages of the passive time reversal does not show to the higher order modulation such as the16QAM modulation, so the lower order modulationexperimental data in channel-tank remains to select and analyze sooner.