Research On Key Technologies Of Self-interference Cancellation For In-band Full-duplex Underwater Acoustic Communication

In-Band Full-Duplex Underwater Acoustic(IBFD-UWA)communication technology can transmit and receive communication signals in the same frequency band at the same time.Theoretically,its frequency utilization efficiency can reach twice of the traditional half-duplex UWA communication.Therefore,it holds great research significance and application value for underwater acoustic communication where available spectrum resources are seriously limited.The main challenge in the implementation of IBFD-UWA communication is to cancel the local Self-Interference(SI)signal.This thesis focuses on the problems that SI signal composition and channel structure of IBFD-UWA communication modem in practical engineering application are not clear,the performance of analog Self-Interference Cancellation(SIC)is affected by hardware parameters,and the performance of channel estimation and digital SIC is limited due to the time-varying nature of SI propagation channel.Based on the radio Full Duplex(FD)communication technology,the theoretical research framework of SI channel estimation and SIC for IBFD-UWA communication system is constructed,and the preliminary exploratory study is carried out based on this framework.The proposed framework consists of propagation domain SI channel modeling technology,Digitally Assisted Analog Self-Interference Cancellation(DAA-SIC),digital SI channel estimation and SIC technology for time-varying shallow water channel.In view of the lack of research on the Self-Loop Interference(SLI)component propagation channel and the complexity of the measured Self-Multipath Interference(SMI)channel in the propagation domain SI channel modeling,the corresponding propagation domain SI channel modeling has been done for SLI and SMI respectively.A model for SLI propagation process based on finite element analysis is proposed and the SLI propagation Channel Impulse Response(CIR)is obtained according to the simulation results.The SMI channel is modeled by ray acoustic theory,and the time-varying SMI channel model is derived by combining the characteristics of wind-induced sea surface fluctuation and statistical channel model.The performance of the proposed SLI and SMI channel modeling methods are verified by simulation results as well as lake tests.The research results show that the SI propagation channel has obvious clustering characteristics,and provide a more practical basis for the follow-up study of channel assumption in the process of SIC in analog domain and digital domain.Based on the simulation and experimental results,several recommendations are proposed to reduce the SLI in IBFD-UWA communication system for shallow water engineering application.Considering the fact that the performance of analog domain SIC is affected by hardware parameters(e.g.effective quantization bits of analog-to-digital converter,nonlinear distortion of power amplifier),the research is carried out according to this route:demand analysis of SIC,analysis of influencing factors of hardware parameters and solutions.Firstly,combined with the passive sonar equation and acoustic propagation theory,the total demand of SIC under certain marine environment conditions and different communication distances is given,and combined with the influence of Analog-to-Digital Converter(ADC)quantization bits at near-end receiver.The theoretical and simulation analysis of the performance of the basic scheme of analog domain SIC is completed,and the reasons for the limited performance of analog domain SIC are given.According to the measured results of SI propagation channel characteristics,DAA-SIC technology is introduced to solve the shortcomings of the analog domain SIC scheme based on limited number of taps so that it can meet the requirements of IBFD-UWA communication system.A PA-DAA-SIC scheme based on Power Amplifier(PA)output signal replica and the Memory Polynomial(MP)-DAA-SIC scheme of PA output reconstruction based on the MP model of Volterra-series are analyzed theoretically and simulated.It is concluded that the SIC performance of the above two schemes in practical application is limited by the effective quantization bits of the auxiliary link and the PA noise respectively.Based on the above conclusions,the DPD-MP\PA-DAA-SIC scheme based on Digital Pre-Distortion(DPD)and the PANDA-DAA-SIC scheme based on Power Amplifier Noise Data Acquisition(PANDA)are proposed respectively.The performance of the two new DAA-SIC schemes are verified by theoretical simulation and circuit simulation under measured hardware parameters.The results show that the DPD-MP-DAA-SIC scheme has relatively stable performance under the condition of fluctuation of effective quantization bits of auxiliary link.Considering the influence of the transmitter noise and the effective quantization bits of the auxiliary link,the PANDA-DAA-SIC scheme has the optimal SIC performance in the analog domain under the above conditions.Aiming at the performance limitation of channel estimation and digital domain SIC in time-varying SI propagation channel,two time-varying SI channel estimation and SIC techniques are proposed respectively from the perspective of time-varying channel tracking and the performance improvement of adaptive filter under the influence of time-varying channel,two time-varying SI channel estimation and cancellation techniques are proposed.Firstly,the performance of time-invariant and time-varying SI channel estimation is studied and analyzed using the Recursive Least Squares(RLS)adaptive filter.Then,based on the modeling results of time-varying SI channel,the characteristics of time-varying SI propagation channel are analyzed in combination with IBFD-UWA communication node layout scenario in a practical engineering application,and its local stability features are obtained.Based on this features,a channel tracking method,driven by clustering path change feature with Kalman filter and RLS filter as the core,is proposed.By tracking the channel delay and amplitude of the strongest tap of each changing clustering,the time-varying SI channel structure can be updated.From the perspective of Variable forgetting factor(VFF),an Improved VFF-RLS(IVFF-RLS)filter based on three-stage forgetting factor change rule is proposed for IBFD-UWA communication system,which is based on the overlapping characteristics of SI signal and expected signal and the prior estimation error component analysis of RLS filter.The instantaneous prior estimation error in RLS iterative process is stabilized by M-estimation theory.The signal overlap state is judged by stabilizing the prior estimation error.On this basis,the change of forgetting factor is controlled by combining the prior energy intensity information of expected signal.Finally,the simulation and result analysis of conventional RLS,VFF-RLS and IVFF-RLS are carried out based on channel estimation accuracy,SIC performance and Bit Error Rate(BER)as performance indices,and the effectiveness of IVFF-RLS filter is further established by lake measured data.Results show that IVFF-RLS filter has the best performance of the above three schemes under BER index.