Research On Signal Processing Method For High Frequency Radar Based On Sparse Stepped Frequency Signal

High frequency radar is a kind of over the horizon radar,which works in short band from 3MHz to 30 MHz.However,a large number of electromagnetic interference in this frequency band result in narrow continuous frequency band,which limits the bandwidth expansion of radar transmitting signal and further affects the distinction ability of radar.In order to ensure that the radar still make full use of frequency resources in the discontinuous frequency spectrum,the synthetic bandwidth signal gradually shows its unique advantages.As a typical synthetic bandwidth signal,stepped frequency signal has been studied for many years in synthetic aperture radar and various imaging radars.It has gained a lot of experience in theory and technology,but there are still few related researches in the field of high frequency radar.Because of the difference between HF radar and other radar systems,the existing methods and conclusions can not be directly transplanted to HF radar.Therefore,according to the characteristics of HF radar,based on the existing stepped frequency signal,an improved sparse stepped frequency signal is proposed.The corresponding target parameter estimation method is also studied,so that HF radar can better adapt to the complex environment,more flexiblility and accuracy in locating targets.It has important application value and broad development prospects both in military and civil fields.This paper deeply analyzes the special electromagnetic environment of HF radar and also the principle of stepped frequency signal.On this basis,the research work in this field is systematically carried out from the design of stepped frequency signal in sparse spectrum,the range-Doppler estimation of target and the range angle decoupling of frequency diversity array.Effective solutions are proposed for several key problems.The specific contents of each part are summarized as follows:For the design of stepped frequency signal in sparse frequency spectrum,a new signal called sparse stepped frequency signal is proposed to adapt to the discontinuous spectrum environment for HF radar.On the basis of summarizing the rule of actual electromagnetic spectrum,this paper compares the time-domain waveform and ambiguity function performance of commonly used FMCW,FMICW and phase coded signal in HF radar.At the same time,the fundamental principles and properties of stepped frequency pulse signal are studied,and the CRLB performance of multiple parameters joint estimation is derived.The influence of discontinuous spectrum on the estimation performance is analyzed.According to the spectrum constraints and signal characteristics,a sparse stepped frequency signal with higher flexibility,stronger anti-interference ability and better adaptability is designed.To further improve the waveform parameters,a joint optimization method of frequency codeword and sensing matrix is proposed.As an important parameter in the process of compressed sensing processing,the sensing matrix in the parameter design method provides the minimum criterion of correlation function.Through continuous iterative updating,the optimized sparse frequency coding is obtained,and then the corresponding optimized waveform with spectrum sensing and environment adaptability is constructed.A signal model based on compressed sensing method is established to estimate the range-Doppler of targets.At the same time,several critical problems of the estimation processing are studied with the proposed processing methods.In the process of target range-Doppler estimation for sparse stepped frequency signal,due to the mutual restriction between the signal parameters,direct estimation usually leads to the problem of distance mismatch.Therefore,an effective information extraction method is proposed.This method can not only avoid the loss of target information during sampling,but also obtain the accurate information corresponding to the real target position through the extraction of effective points,so as to eliminate the error caused by distance mismatch.The velocity of the target also affects the range estimation,which makes the range spectrum shift,widen and moving.Therefore,the impact of velocity is also discussed in detail,and a multi-objective velocity compensation method is proposed to effectively eliminate the estimation error caused by velocity.Besides,when the target falls out of grid,it can not get the accurate information of targets because the compressed sensing method is applicable to the target on the grid.Therefore,this paper proposes one dimension and two dimension sparse target estimation methods based on matrix completion,which can break through the grid limitation and further improve the accuracy of target parameter estimation.The range-angle decoupling in frequency diversity array which composed of step frequency signal and linear array is studied.A method of uncoupled non-uniform stepped frequency optimization is proposed.The effective beam focusing can be obtained by using this method.The frequency diversity array radar is composed of step frequency signal and linear array.This kind of FDA is based on the traditional phased array,which distributes each single frequency signal in the stepped frequency group to every antenna element.There are not only phase shift but also equispaced frequency shift between the elements.In view of the coupling phenomenon in the target angle estimation,a new nonlinear frequency offset combination design scheme is proposed under the condition of continuous spectrum and sparse spectrum respectively.Especially the sparse stepped frequency optimal allocation is able to adapted to the special environment of high frequency.It can eliminate the influence of coupling on the angle estimation of the target,and make the generated beamforming an effective focus to the designated position.Finally,the feasibility of the proposed sparse stepped frequency signal waveform optimization and parameter estimation method in HF radar target detection are verified by computer simulation.