Frequency Estimation Algorithms Based On Newton Iteration And LS Linear Regression And Their Underwater Applications

With the development of marine resources,underwater frequency estimation technology is widely used in many fields,such as flow velocity measurement and underwater positioning and navigation.Underwater acoustic channel is complex and changeable,with narrow bandwidth and serious noise interference,which makes the received signal distorted.Fast Fourier transform(FFT)has gradually become one of the most popular methods in the field of frequency estimation due to its fast results and intuitive physical meaning.However,the problems of spectrum leakage and fence effect in FFT technology directly affect the accuracy of frequency estimation.In this paper,the frequency measurement of single frequency signal is studied based on FFT.In order to achieve accurate frequency estimation of underwater acoustic signal,the following research is carried out in this paper:(1)Research on underwater acoustic frequency estimation algorithm based on Newton iteration.The filter is used to roughly eliminate the multi-path interference with different Doppler frequency offset,and then the FFT is used to roughly estimate the frequency of the direct wave.The rough estimation value is used as the initial value of the first-order Newton iteration for precise frequency estimation,which is iterated twice to greatly compensate for the frequency estimation error caused by the fence effect.And finally achieve accurate underwater acoustic frequency estimation.Through simulation tests,the performance of the proposed algorithm is compared with many algorithms to verify the effectiveness of the proposed algorithm.(2)Research on underwater acoustic frequency estimation algorithm based on Least Squares(LS)linear regression.The rough frequency estimation of the received signal is realized based on FFT,the frequency difference equation is obtained,and then the windowing sum is carried out to improve its signal-to-noise ratio.Then the additive noise is equivalent to phase noise,and then the error of the rough frequency estimation is obtained by linear regression of the instantaneous phase of the signal;Add the rough frequency estimation value and the rough estimation error to obtain the accurate frequency estimation.Through simulation tests,the performance of the proposed algorithm is compared with many algorithms to verify the effectiveness of the proposed algorithm.(3)Carry out actual experiments in multiple waters.For the underwater acoustic frequency estimation algorithm based on Newton’s iteration,the water tank,pool and Jiaozhou Bay were experimented,and the underwater acoustic frequency estimation algorithm based on LS linear regression was experimented in the water tank and Jiaozhou Bay under ship noise.The experimental results verify the good performance of the two algorithms under low signal-to-noise ratio,and show that the two algorithms have good application value in the fields of water flow rate measurement and underwater positioning and navigation.(4)Research on the realization of underwater speed measurement module based on Digital Signal Processor(DSP).The proposed two algorithms are embedded into DSP respectively.The underwater velocity measurement module based on TMS320C6657 hardware platform is realized.The data of Jiaozhou Bay floating experiment are used to verify the effectiveness of the module.Based on FFT,this paper proposes a low complexity underwater acoustic frequency estimation algorithm based on Newton iteration and a high-precision underwater acoustic frequency estimation algorithm based on LS linear regression.Newton iteration and LS linear regression techniques are used to compensate for the estimation error caused by the fence effect in the FFT spectrum respectively.It ensures the effectiveness of frequency estimation.The effectiveness and stability of the two algorithms are verified by theoretical analysis,simulation experiments and measured results in multiple waters.The implementation methods of the proposed two frequency estimation algorithms based on DSP chip underwater velocity measurement module are given,and the module is tested to verify the effectiveness of underwater velocity measurement module.