| In recent years,with the rapid development of Unmanned underwater vehicle(UUV),the application of underwater reconnaissance technology based on unmanned platform is more and more extensive.The detection of non-cooperative underwater acoustic communication signals needs to go through the process of modulation identification,parameter estimation,demodulation and decoding,etc.After the signal has been modulated identification,the parameters of the signal need to be accurately estimated,and the estimation of carrier frequency parameters is an important topic.In non-cooperative scenarios,all prior information of underwater acoustic communication signal is unknown.Due to the diversity of underwater acoustic communication signal forms and the doppler caused by the maneuvering motion of UUV,the accuracy of parameter estimation method based on Fourier transform is low.Therefore,it is very important to select the appropriate algorithm to realize the high-precision parameter estimation of the signal.This paper focuses on the parameter estimation methods of non-cooperative underwater acoustic communication signals.The main contents include typical underwater acoustic communication signal characteristics and basic time-frequency analysis theory research,parameter estimation methods based on second-order time-frequency analysis and third-order time-frequency analysis,etc.At the same time,theoretical analysis and simulation verification of the above two types of parameter estimation algorithms are carried out.The practicability of these two kinds of parameter estimation methods is verified by measured data and semi-measured data.For the commonly used forms of underwater acoustic communication between noncooperating parties,this paper lists four typical underwater acoustic communication signals,including Linear Frequency Modulated signal that exists as the synchronization head of underwater acoustic communication,the Binary Phase Shift Keying signal and the Binary Frequency Shift Keying signal as the typical communication signal and the Direct Sequence Spread Spectrum signal,and then analyzes the time domain and frequency domain characteristics of signals.The research on parameter estimation of non-cooperative underwater acoustic communication signal mainly consists of two parts.First,the research on the parameter estimation method based on the second-order time-frequency analysis is mainly composed of four parts: 1.The research is carried out around short-time Fourier transform algorithm,and the principle and performance characteristics of the algorithm are simulated and analyzed.Due to its poor time-frequency aggregation,a synchronous compression transform algorithm based on short-time Fourier transform and a local maximum synchronous compression transform algorithm based on short-time Fourier transform are introduced to improve the time efficiency and frequency aggregation of the algorithm.Then,the parameter estimation performance of the algorithm is simulated and analyzed by four typical underwater acoustic communication signals.2.The principle and performance characteristics of Wigner-Viller distribution(WVD)algorithm are simulated and analyzed.In view of the cross term brought by the nonlinearity of WVD,and then affect the accuracy of parameter estimation,this shortcoming.The Smooth Pseudo Wigner-Viller distribution(SPWVD)algorithm is introduced to reduce the influence of cross terms on the accuracy of parameter estimation.Then,the parameter estimation performance of the algorithm is simulated and analyzed by four typical underwater acoustic communication signals.3.The research is carried out around the fractional Fourier transform algorithm,and the principle and performance characteristics of the algorithm are simulated.Then,the parameter estimation performance of each algorithm is analyzed through typical underwater acoustic communication signals.4.The parameter estimation error of the above algorithms is compared and analyzed,and the difference of the estimation accuracy of the parameters of each algorithm is compared.The Fractional Fourier Transform has the highest estimation accuracy of the carrier frequency parameters of simulation signals in this paper,and the results are verified by the measured data.Secondly,under non-cooperative conditions,due to the broadening of doppler spectrum caused by UUV motion,non-cooperative signals will be distorted into multi-component highorder phase signals.Therefore,this paper introduces a parameter estimation method based on third-order time-frequency analysis.It includes a parameter estimation method based on RadonCPF transform,a parameter estimation algorithm based on CCD transform and a parameter estimation algorithm based on RCCD transform.The algorithm principle and characteristics of each algorithm is introduced.Finally,the robustness of the three algorithms is analyzed,and the reasons for the performance are pointed out.The difference of parameter estimation error of each algorithm is simulated and analyzed under the condition of low signal-to-noise ratio and UUV maneuvering motion.At the same time,the semi-measured signal is used to verify that the introduced three algorithms can improve the parameter estimation performance of noncooperative underwater acoustic communication signals based on UUV,and the RCCD algorithm has the highest parameter estimation accuracy for the signal in this paper. |
