Investigation On Receiver Systems Of Frequency Diverse Array Radar And Its Parameter Estimation

Frequency diverse array(FDA)is a new technique.The FDA radar has drawn much attention in recent years.Compared with traditional phased-array,FDA introduces frequency increments on the common carrier frequency across its array elements to produce a far-filed transmit beampattern that simultaneously depends on the time,angle and range,which brings FDA additional degrees-of-freedom(DOFs)and many promising applications.Current FDA researches focus on signal transmission,while little work has been done about effective receiver design for FDA transmitted signals.Therefore,starting from the FDA receiver design,we aim at exploring reasonable and effective receiver design for FDA signals,and meanwhile study the FDA radar related parameter estimation issues.In summary,the research contributions are listed as follows:1.Effective FDA radar receiver processing together with target parameters estimation is studied:Firstly,three types of existing FDA receivers are comparatively analyzed in noise environment,where both overlapping and non-overlapping FDA signals are considered.It is proved that the hybrid architecture is optimal for receiving non-overlapping FDA signals,while the full-band architecture followed by discrete Fourier transform(DFT)performs better for overlapping FDA signals.Next,in order to fully exploit coherent array gain of FDA transmitted signal,a frequency-domain signal model is proposed,which enables more direct and fairer comparisons of transmit coherent gain and localization performance between FDA and phased-array(PA)radar.With this model,it is found that FDA has better ranging performance and PA has better angle estimation performance.Finally,since the advantages of FDA returned signals cannot be fully exploited by traditional matched filters,a receiver structure is proposed for coherent pulsed-FDA,where the FDA signals are processed by multi-channel mixers followed by matched filters.The validity of the proposed receiver is verified by both theoretical analysis and simulation results.Because of the fact that the coherent pulsed-FDA receiver can handle both overlapping and non-overlapping FDA signals,this receiver structure is meaningful in practical applications.2.A fast parameter estimation algorithm for monostatic FDA-MIMO radar is proposed:By utilizing the waveform orthogonality,the FDA transmitted signals can be separated by the receiver array.Because of the complexity of transmit steering vector,existing methods usually involve complex two-dimensional(2D)searching to jointly estimate the angle and range of targets,which is too computational complexity and time-consuming to implement.To improve the computational efficiency,a fast two-stage estimation algorithm is proposed for the FDA-MIMO radar,which is achieved by converting the 2D estimate to several single-dimensional estimators which can be easily resolved by conventional spectral estimation techniques.It is shown that the proposed algorithm can be applied to FDA-MIMO models with both small-scale and large-scale arrays at a low complexity.Meanwhile,for moving targets,an unstructured maximum likelihood(UML)method is proposed for fast estimating the 3D parameters including the Doppler frequency,angle and range.Compared with traditional ML methods,the UML method has lower computational complexity and is easier to implement.Theoretical analysis shows that the UML method would lead to inherent performance loss for angle estimation,which also applies to other two-stage estimation methods.3.Two search-free methods are proposed for bi-static FDA-MIMO radar target parameters estimation.Referring to bi-static FDA-MIMO radar target parameters estimation,it is pointed out that the direction-of-arrival(DOA)and departure-of-arrival(DOD)are not identical anymore,and consequently the coupling phenomenon between the DOD and the range of a target cannot be solved by a linear array receiver.Non-linear frequency increments should be adopted in this case.Moreover,the parameters to be estimated become 3D,including DOD,DOA and range,which means quite high computational complexity if 3D searching algorithms are adopted.To solve this problem,ESPRIT-based and PARAFAC-based algorithms are proposed for the FDA radar using sub-array configuration and random frequency increments,respectively.In addition,to address FDA periodical phase ambiguity,a resolving method is proposed for such search-free algorithms.Theoretical analysis and simulation examples show that the proposed search-free methods have comparatively good precision of estimation while remain low complexity.The ESPRIT-based algorithm is more suitable to a small-scale array probing multiple targets,while the PARAFAC-based one performs better for a large-scale array probing only a few targets.4.One passive receiver design and related parameter estimation algorithm are proposed for FDA radar.Considering that the flexible transmit configurations of FDA radar,and the transmit signal spectra may discretely distribute on rather wide frequency band,which results in requirement of quite large sampling rate,an under-sampling array structure based on modulated wide-band converter(MWC)is proposed to achieve passive reception of unknown FDA signal with under-sampling rate.Moreover,an MWC-based joint estimating method of signal frequency and DOA,followed by the corresponding pair-matched methods,is proposed to solve the possible under-sampling problems.The feasibility and stability of the proposed structure and estimating methods are verified by simulation examples.