Research On Adaptive Underwater Acoustic Downlink OFDMA Communications Under Time-varying Channel

The complex underwater acoustic(UWA)channel has the characteristics of time-spacefrequency variation,which takes great challenges to the design of UWA communication networks.Considering the robustness of UWA communication and networking system in complex time-varying channels,aiming at the large propagation delay,frequency selective fading,time-varying characteristics,limited bandwidth and energy of the UWA channel,this paper combines adaptive communication technology to improve the reliability and effectiveness of the UWA systems.Due to the time-varying and large transmission delay of UWA channel,as well as the inevitable errors caused by the signal processing technology,the feedback channel state information(CSI)in adaptive UWA communication is usually imperfect.Therefore,this paper focuses on the multi-user access method of adaptive UWA communication network under time-varying UWA channel: considering cross layer design,which combines with the physical characteristics of UWA channel and network layer resource scheduling,on the basis of fully studying the characteristics of time-varying UWA channel,the communication parameters in UWA network are dynamically adjusted based on imperfect CSI;Through channel optimization and channel prediction,the impact of inaccurate CSI on the overall performance of the network is reduced,so as to improve the overall rate of the network and ensure its reliability.The main research contents of this paper are as follows:This paper analyzes several typical propagation characteristics of UWA channel and its impact on UWA communication and networking system,for providing theoretical support for UWA channel modeling.A time-varying UWA channel modeling method is proposed.The channel fading is divided into large-scale fading and small-scale fading.The random component introduced by scattering is considered.At the same time,the motion causing Doppler frequency shift is modeled as stable motion component and random motion component,so that the channel model is closer to the actual UWA channel.Combined with the measured data,the characteristics of underwater acoustic time-varying channel are analyzed.This paper constructs a UWA downlink adaptive orthogonal frequency division multiple access(OFDMA)system by combining orthogonal frequency division multiplexing(OFDM)technology and adaptive communication technology,using OFDMA as the multiple access mode of UWA network.Including: constructing the physical layer modulation and demodulation method and the CSI reconstruction scheme based on data fitting,and establishing the data link layer adaptive resource allocation model.The carrier-bit-power joint adaptive resource allocation algorithm is proposed.According to the feedback CSI,the UWA channel adaptive OFDMA system jointly optimizes the carrier,bit,power,and other parameters of multiple nodes,and realizes parallel conflict free multiple access through OFDM.Theoretical analysis and simulation show that adaptive carrier-bit-power allocation can adaptively adjust the communication parameters of each node,not only ensure the robustness of a single node,but also make the system obtain multi-user gain and optimize the overall performance of the network.The CSI reconstruction scheme based on data fitting can reduce the amount of feedback information and maintain the effectiveness of CSI.This paper studies the influence of outdated CSI on UWA adaptive communication network system,and proposes a low overhead adaptive downlink OFDMA system with low overhead under time-varying UWA channel.The channel correlation attenuation factor(CCA)and longterm statistical of CSI are defined as two impact factors to describe the outdated CSI,for optimization.A CCA based CSI selection approach is proposed.According to the CCA factor,the CSI for feedback is selected between the outdated-instantaneous CSI and the outdatedaverage CSI.The measured data processing results show that the CCA factor is related to the delay,and the long-term statistics of CSI are stable,which means that the outdated-average CSI can be used for channel feedback.Experimental results show that the CCA based CSI selection approach has better performance than only using outdated-instantaneous CSI or outdatedaverage CSI for feedback.Aiming at the complex correlation of UWA channel in time domain and frequency domain,on the premise of collecting enough historical channel observations,an effective online prediction model of UWA channel is proposed.Including: a time-frequency domain joint channel prediction model(TFPre Net)is designed,which combines one-dimensional convolutional neural network(1D-CNN)and long-term short-term memory neural network(LSTM)to predict the CSI of UWA channel in time-frequency domain.In order to verify the effectiveness of the prediction,the channel prediction algorithm is also applied to the UWA downlink OFDMA system.The measured UWA channel is used in the OFDMA system,and the joint subcarrier-bit-power adaptive resource allocation scheme is adopted.In order to further improve the performance,an off-line on-line prediction mechanism is proposed to make the continuous prediction results more stable and effective.Simulation and experimental results show that the prediction model based on TFPre Net can effectively predict the future CSI,and its performance is better than the existing recursive least squares(RLS),Kernel adaptive filter(KAF)predictors,and common DNN based predictors.Aiming at the problem that the feedback CSI obtained by the central node in UWA adaptive communication and networking system is usually imperfect,based on the limited historical channel observation,considering the two factors of CSI outdated and channel estimation error,an inaccurate CSI optimization method based on Gaussian Markov model and per-subcarrier channel time correlation factor(PSCTC)is proposed.The scheme is applied to the downlink UWA OFDMA multi-node communication system.In this method,PSCTC factor is used weigh the errors caused by channel outdated and estimation error.Simulation and experimental results show that the optimized channel can effectively improve the communication quality in timevarying UWA channel.