Research On Cognition Based Underwater Transform Domain Communication System

Underwater acoustic communication network,as the most important underwater infrastructure,is playing an irreplaceable role in sea environment sensing,underwater surveillance and so many other missions in both military and civilian purpose.However,due to the limited frequency band of underwater acoustic channel,the underwater acoustic communication system is vulnerable to be interfered by other underwater acoustic systems including marine mammal activities.At the same time,because the multipath of underwater acoustic channel is long and complex,it will cause serious frequency selective fading to the underwater acoustic communication signal.These factors make great obstacle of the development of high speed and intelligent underwater acoustic communication systems.In order to 1)utilize the valuable frequency resource of underwater acoustic channel more efficiently,2)decrease interference from human and natural activity,in the other hand,reduce the intrusion of human activities on the nature,and 3)enhance the ability to resist the multipath loss,this paper proposes a cognitive based underwater acoustic transform domain communication system?CU-TDCS?.Cognitive underwater communication systems build up rules of modulation method and receiver algorithm over time through learning from continuous experiential interactions with the underwater environment,and thereby deal with channel uncertainties and possible adversary jamming by making corresponding changes in certain operating parameters?e.g.,transmit-power,carrier-frequency,and modulation strategy?in real-time.The waveform of a traditional communication system is synthesized in the time domain according to the frequency band assigned to a specific user.Different from that,the proposed CU-TDCS synthesizes the communication spectrum after eliminating the interfered band in the frequency domain or in the wavelet domain,and then get the time domain waveform needed by the transmitter through the corresponding transformation,so as to avoid the interference.Cognitive technology is the basis and an undivided part of TDCS.Though TDCS can utilize Fourier,Wavelet and Fractional Fourier transform as the transform domain converter,but in this paper,we only discuss the Fourier transform,which is easier to realized and more efficient in TDCS.The CU-TDCS is still in the initial stage in the world,and it is in the ascendant.Most of the study of this area is focus on discovery the possibility to build a cognitive system,but the underwater TDCS is not addressed in literature.The TDCS in underwater acoustic environment has not yet been described in the open literature.First of all,a system model of CU-TDCS is proposed,and cognitive theory and underwater acoustic transform domain communication technology are introduced into the field of underwater acoustic communication.Usually,the traditional underwater acoustic communication systems only use complex signal processing algorithms at the receiver end to reduce interference on decoding.But different from that,the CU-TDCS analysis real-time interference of environment cognitively in the receiver end,and sent to the transmitter through the feedback channel.So that,the transmitter and receiver are composed of a closed-loop feedback system,and adjust system parameters adaptively according to the environment and interference.The system is adaptive not only at the receiver,but also at the transmitter.After propose the physical and logical system model,the work flow of the system and the cognitive process is carried out.To solve the problem of false alarm and miss report in signal detection,a new method for spectrum sensing is carried out.In order to complete multi-user communication and cognition information transmitted through the feedback channel,A briefly design of underwater acoustic communication protocol of cognitive network is presented.In order to further improve the channel bandwidth efficiency,a new modulation technology combined orthogonal frequency division multiplexing?OFDM?with TDCS which usually use cyclic shift keying?CSK?called cognitive OFDM?C-OFDM?is proposed.Based on the cognition of the underwater acoustic environment,C-OFDM eliminates the sub carriers which are interfered and the sub channel is deep fading in frequency domain,which means,only the passbands of the underwater channel are used to carry the transmission information.It can effectively improve the communication rate,on condition of the reliability of information transmission.Simulation results show that,in low SNR?SNR=0dB?and complex multipath environment,C-OFDM based CU-TDCS can effectively eliminate the stopband,and once the threshold is set properly,the bit error rate can reach 0.The accurate estimation of the band range of interference signal is the precondition of eliminating the interfered channel,and it is also a major work of underwater acoustic cognition.Therefore,the detection and parameter estimation techniques of underwater acoustic interference signal are studied in this paper.After analyzed the modulation mode and frame structure of commercial communication sonar commonly used,the signal structure of highly possible interference is acquired.And this signal is tested by the algorithm for parameter estimation in this paper by simulation and experiment.In order to estimate the parameter of interference signal,?1?in Gaussian noise environment,we carry out iterative adaptive approach spectrum estimation?IAA?and sparse covariance-based estimation?SPICE?,and in impulse noise environment we introduced l₁-IAA,l_p-IAA and l₁-SPICE,l_p-SPICE to estimate frequency and time slot.The simulation and experimental data indicates that IAA and SPICE algorithm can achieve higher spectrum and time resolution of frequency hopping signal,compared with the classical periodogram method.In the sea experiment,l₁-IAA,l_p-IAA,l₁-SPICE spectrum and l_p-SPICE estimation technique has good ability of impulse noise.In addition,according to the state of the art of underwater acoustic communication network,multi carrier communication mode is a common modulation method.A detection and parameter estimation method based on cyclic prefix is carried out.Theoretical and sea experimental indicate that the carrier interval estimation accuracy of multi carrier frequency hopping underwater acoustic communication signal is 0.014%of the sub carrier bandwidth.Finally,combined the underwater cognitive technology with TDCS,we propose 3 new underwater channel estimation method in CU-TDCS simulation model based on CSK modulation for the first time in underwater acoustic area.Due to the sparse characteristics of underwater acoustic channels,3 low complexity sparse channel estimation techniques based on Dichotomous Coordinate Descent?DCD?iteration technology are proposed in CU-TDCS framework.1)Aiming at the problem of error transfer in conventional orthogonal matching pursuit?OMP?algorithm for sparse underwater acoustic channel estimation,we develops a Multipath Matching Pursuit?MMP-DCD?sparse underwater acoustic channel estimation method with DCD;2)In order to eliminate the error accumulation from one path matching pursuit and reduce the computational burden of Orthogonal Least Squares?OLS?as well as in other greedy algorithm,a DCD based Multiple Orthogonal Least Square?MOLS-DCD?is presented;3)for the complicated requirements that conventional sparse channel estimation method based on compressed sensing must have the knowledge of sparsity of the channel or noise variance to stop iterating,e.g.greedy algorithm,which is not practical in most circumstance,we developed Multipath Orthogonal Least Square?MOLS?estimation based on cross validation?Cross Validation CV?.Combining the interference band estimation technology and the proposed sparse channel estimation algorithm,we build the underwater acoustic TDCS system based on the measured acoustic velocity profiles and bottom topography of Songhua Lake.a TDCS communication signal with waveform level is generated by using hi fidelity Waymark virtual underwater acoustic signal simulator.The proposed sparse channel estimation algorithm and the underwater acoustic TDCS scheme based on sparse channel estimation algorithm are evaluated under three typical launch and reception configuration conditions.By means of simulation,we get the results that?1?Utilizing the DCD iteration,the computation burden of Least Square?LS?algorithm in MMP-DCD,MOLS-DCD and MOLS-DCD-CV has reduced from O?K²?to O?K?,where K is the length of the LS;?2?in sparse underwater acoustic channel conditions,the MMP-DCD,MOLS-DCD and MOLS-DCD-CV algorithm have better performance compared to the conventional LS estimation algorithm;?3?MOLS-DCD algorithm with CV,can achieve a better performance than the MOLS algorithm with known sparsity of the channel,even MOLS-DCD-CV is without the knowledge of sparsity information;?4?Compared with the TDCS using LS channel estimation algorithm,the TDCS system based on sparse channel estimation algorithms proposed in this paper,has better performance in BER sense.In this paper,the composition of the system,system processes procedure,and cognition to the environment,and some other key technologies of CU-TDCS are studied and proved,which laid the technology foundation for future cognitive underwater communication networks and underwater covert communication.The research results in this paper can be applied not only to the field of underwater acoustic communication,but also to the field of underwater detection,which can be applied to the construction of cognitive sonar system.