Font Size: a A A

Sinusoidal Frequency Modulation Fourier Transform And Research On Micro-doppler Signature Retrieval For Radar Targets

Posted on:2015-02-17Degree:DoctorType:Dissertation
Country:ChinaCandidate:B PengFull Text:PDF
GTID:1108330509461032Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
The motion in addition to the bulk motion, made by the object or any structural component of the object, is called the micro motion, which is always weak and oscillatory. The micro motion provides a vital signature to the target, and plays an unique and important role in the radar-based target recognition and imaging areas. The micro motion parameters estimation is now facing some technical bottlenecks, i.e., the precision of the traditional joint time-frequency analysis(JTFA) based methods is too low, while the model-based methods have strict limitation on the micro motion form. This has brought serious challenges to the present micro parameters estimation theory, which is mainly based on the above two kinds of methods. This dissertation investigates on the retrieval of the micro motion signature of radar target, firstly defined the sinusoidal frequency modulated signal space and put forward a series of micro motion parameters estimation methods, such as the sinusoidal frequency modulation Fourier transform. These methods greatly improves the precision of micro motion parameter estimation, and extend the applicable scope of the micro motion model, which can be applied to the retrieval of the micro motion signature and radar target imaging.Chapter 1, the introduction, illustrates the background and significance of this research, reviews the development and the current techniques in the micro motion signature retrieval, and introduces the main content of this dissertation.Chapter 2 embarks from the model of micro motion, which is divided into the harmonic and complex micro motion, the common and micro-range micro motion. And then,the radar echo from the micro motion target is modeled based on the multi-sinusoidal frequency modulated signal, which can include all the above micro motion forms. Next, the limitations of the JTFA-based method and the model-based methods are studied. Taking the pseudo wigner distribution as an example, quantitative analysis of the bias and random error in the instantaneous frequency estimation of sinusoidal frequency modulated signal based on the JTFA-based method was studied, which theoretically reveals the bottleneck of the JTFA-based method in the analyzing the sinusoidal frequency modulated signal.Chapter 3 establishes the sinusoidal frequency modulated signal space, and then, puts forward the sinusoidal frequency modulation Fourier transform(SFMFT). Unlike the JTFA which apply a sliding short-time window to perform an instantaneous approximation, the SFMFT accumulates the micro-doppler modulation information for a long time,and consequently improves the precision of estimation significantly. Moreover, from the perspective of the signal model, the SFMFT solve the problems of multitudinous parameters and complicate resolving process in the multi-sinusoidal frequency modulated signal analysis radically. The research on the precession and the estimable range of the micro motion amplitude is given after the proposal of SFMFT. We puts forward the concept of micro-Doppler to noise ratio, analyzes the problem of phase ambiguity and the phase unwrapping process in SFMFT, and then studies the SFMFT of multi-component multisinusoidal frequency modulated signal. According to the above analysis, the experiential formula of the normalized root mean square error in the analysis of sinusoidal frequency modulated signal, as well as the experiential formula of SFMFT results towards the multi-component multi-sinusoidal frequency modulated signal, are summarized. Next, two kinds of corresponding algorithm to solve the concrete issues have been brought out and studied, i.e., the instantaneous frequency estimation based on SFMFT and vehicle vibration spectrum estimation based on SFMFT. Theoretical analysis and simulation experiment results showed that, the SFMFT can achieve the precise estimation of weakmodulated multi-sinusoidal frequency modulated signal, improve the estimation precision and antinoise performance significantly compared with the existing methods, lower down the estimable micro motion amplitude threshold, and retrieve the micro motion signature of the targets with unknown micro motion form.Chapter 4 puts forward the sinusoidal frequency modulation sparse recovery(SFMSR) algorithm, which construct the Fourier modulation dictionary and introduce the sparse bayesian learning(SBL) to the sinusoidal frequency modulated signal space. The SFMSR can achieve a precise analysis for sinusoidal frequency modulation under the condition of low data rate, non-uniform sampling. Compared with the existing methods based on sparse recovery, the SFMSR reduces the problem to one-dimensional parameter optimization, which has a higher robustness, a more precise estimation and a more efficient of estimation process. We also researched on the problem of micro motion spectrum ambiguity, the coherence and the grid size of the Fourier modulation dictionary. Aiming at the precession frequency estimation using low-frequency-band long range radar, which is limited by low data rate, non-uniform sampling and very weak micro-doppler effect,we proposed the SFMSR-based precession frequency estimation method. According to the simulation based on the radar parameters of PAVE PAWS, the proposed method can accurately estimates of the precession frequency through about 5 minutes tracking data accumulation.We propose and elaborate the non-ideal scattering center target imaging method in chapter 5, which can clearly exhibits different types of scattering centers with a large coherent accumulation angle. This imaging method combines SFMFT with HRRP sequences, and in this way to extract the micro motion information of different scattering centers, realize the coarse estimation of the scattering center locations. Next, we establish the composite scattering center model. Through the scattering center type discrimination,sliding-type scattering center parameter estimation based on micro motion spectrum, and the other processes, we realize fine estimation of the scattering center position, and the retrieval of corresponding target’s structure. The proposed imaging method can reflect the spatial distribution of different types target scattering centers, release the information about the size and aspect of the target’s structure. Compared with the traditional imaging method, the proposed method is more intuitionistic and comprehensive.Chapter 6 summarizes this dissertation and discusses the future work.This dissertation achieve a breakthrough in the micro-Doppler signature retrieval from the signal model and the signal processing methods. The research put forward a series of theory and method in the sinusoidal frequency modulated signal space on micro motion parameter estimation, signature retrieval and nonlinear frequency modulated signal analysis, which improves the precision of estimation and expand the scope of the estimable micro motion amplitude significantly. These work will be valuable for the micro motion parameters estimation, radar imaging, target recognition and electronic countermeasures.
Keywords/Search Tags:Micro-Doppler(m-D) Signature, Sinusoidal Frequency Modulation Fourier Transform(SFMFT), Joint Time-Frequency Analysis(JTFA), Sparse Bayesian Learning(SBL), Non-Ideal Scattering Center, Radar Imaging
PDF Full Text Request
Related items