Study On Channel Estimation, Equalization And Power Allocation In Underwater Acoustic Communications

With the rapid development of China's economy,the demands for the exploitation of marine space and marine resources is also greatly growing.As the only effective technical means of remote information transmission under underwater acoustic channel,the strategic significance of the underwater acoustic communication is becoming more and more prominent in the future marine engineering field.However,underwater acoustic channel characteristics include large multipath delay spread and Doppler spread,the severely limited available bandwidth,time-space-frequency variation,which seriously hindered the reliable transmission of high-speed un-derwater acoustic information.In addition,the energy source of underwater acoustic system is very limited because of environmental or practical constraints.This thesis studies on the channel estimation,equalization and power allocation methods in underwater acoustic communications,and uses lake experiment data to verify some methods'performance.The main contributions are listed as follows:1.Aiming to solve the high computational complexity of the conventional time domain multi-channel decision feedback equalizer?McDFE?in large multipath delay spread channel,a pas-sive time reversal based on subarray and multichannel decision feedback equalizer?Sub-PTR-McDFE?method is proposed.The Sub-PTR-McDFE method first divides the receive array into several subarrays,performs passive time reversal processing for each subarray to reduce the subsequent equalization complexity.Then,the output of subarrays is equalized by McDFE to eliminate the residual inter-symbol interference?ISI?,and the computational complexity can be effectively reduced in this way without loss of communication quality.The results of the lake experiment show that the Sub-PTR-McDFE method with two subarrays can achieve the same bit error rate?BER?as McDFE method at the communication distance of 7.4 km and the data rate of 8 kbps,the average BER is 1.6×10^-3,and the computational complexity is 1/5 of McDFE.2.In order to further reduce the BER of Sub-PTR-McDFE method,a passive time reversal based on subarray and bidirectional multichannel decision feedback equalizer?Sub-PTR-BiMcDFE?is proposed.The Sub-PTR-BiMcDFE method combines the Sub-PTR-McDFE method with the bidirectional decision feedback equalization technique,which can reduce the system error rate and improve the communication quality by using the characteristics of low correlation be-tween the two equalizer errors.The results of the lake experiment show that the average BER of the Sub-PTR-BiMcDFE method with two subarrays is 3.4×10^-4at the communication distance of 7.4 km and data rate of 8 kbps.Compared with McDFE method,the BER Sub-PTR-McDFE method is reduced an order of magnitude,and the computational complexity is 1/2 of the McDFE.3.A iterative block decision feedback equalizer with soft detection and iterative channel esti-mation?ICE-SD-IBDFE?method is proposed for single carrier frequency domain equalization underwater acoustic communication system.In this method,the soft information of the equal-izer's output is fed back to further eliminate the ISI.And the iterative channel estimation is adopted in order to deal with the time-varying underwater acoustic channels.The channel es-timation result is weighted to eliminate the error propagation and effectively reduce the BER.Simulation results show that the method is superior to the existing frequency domain equaliza-tion methods under underwater acoustic channel.The results of the lake experiment show that at the communication distance of 1.8 km and 4 kbps data rate,ICD-SD-IBDFE method can im-prove the system reliability compared to the existing frequency domain equalization methods,the single channel unencoded QPSK modulation achieves BER of 10^-3.4.A joint channel estimation and symbol detectionmethod based on the Markov chain Monte Carlo?JCESD-MCMC?method is proposed for the cluster-sparsity characteristics of the under-water acoustic channel.JCESD-MCMC method adopt a modified spike-and-slab prior model in our nonparametric Bayesian learning framework without additional prior information.To avoid deriving a closed-form Bayesian estimate,we apply the Markov chain Monte Carlo?MCMC?technique to achieve joint channel estimation and signal detection.The proposed solution is amenable to being integrated with soft-input soft-output decoding to improve the performance through turbo iteration.Simulation results show that the JCESD-MCMC method has 4 dB per-formance gain compared to the conventional LASSO method at the BER of 10^-4and has good robustness.5.A spatial modulation transmission based on complementary code keying modulation?CCK-SM?and the corresponding receiving method?CCK-IBDFE?are proposed for underwater spa-tial modulation system.The proposed CCK-SM scheme exploits the simple encoding struc-ture of CCK to effectively realize antenna selection and transmitted symbols.At the receiver,the proposed CCK-IBDFE method combines CCK soft demodulation with block iterative deci-sion feedback equalizer to realize an iterative receiver structure to effectively reduce the system BER,which is similar to Turbo equalization.Simulation results show that,with 4 transmit an-tennas and 4 receive antenna MIMO structure,CCK-SM shows nearly 3 dB gain in terms of E_b/N₀at 10^-5BER compared to QPSK-SM scheme.with 4 transmit antennas and 2 receive antenna MIMO structure,the conventional spatial modulation system can not be successfully demodulated,while the CCK-SM system achieves 10^-5BER at 10 dB E_b/N₀,still having good performance.6.In order to make efficient use of the harvested energy in underwater environment,the power allocation technology of underwater acoustic communication system based on energy harvest-ing is studied.A power allocation method based on dynamic programming in underwater en-ergy harvesting communication system?EH-DPPA?is proposed.In this method,the first-order Markov process is used to describe the underwater acoustic channel model and the rate of energy harvesting.The energy consumption of the underwater sensor in the data collection process is considered.Two scenarios are considered for different delay of channel state information?CSI?.The stochastic dynamic programming?DP?approach is employed to derive the optimal alloca-tion policy for both scenarios.A myopic algorithm of dynamic programming?EH-MADPPA?is presented to reduce the inherent computation complexity of EH-DPPA.Our results show the EH-DPPA method achieves substantial performance improvement than Greedy method and EH-MADPPA method provides a good performance-complexity trade-off.Simulation results also demonstrate the performance gain of the proposed method as well as the impact of various prac-tical factors such as average energy harvesting rate,channel steady-state distribution,and CSI accuracy.The work in this thesis would serve as a theoretic base for time domain equalization,frequency domain equalization,sparse channel estimation,spatial modulation and power allocation,and provide some reference for other wirless communication systems.