Research On The Application Of Compressive Sensing In Inverse Synthetic Aperture Radar

Inverse synthetic aperture radar (ISAR) is a kind of high resolution imaging radar. It hasthe capability of getting high resolution images of noncooperative targets in day/nigh,all-weather and long-range conditions. It is with great military and civil values. With thedeveloping of military demands, the traditional ISAR imaging faces following difficulties.1)The traditional imaging methods are faced challenge of higher resolution and real-timeprocessing from imaging objects with more complex movement and higher maneuverability.2)Higher resolution radar imaging with real-time processing results in the increase of stress toradar signal processors, such as A/D, storage, transmission and so on.3) It’s harder to get radarimages with fewer echos or the gap data of echos resulting from the limit of imaging system andhigher maneuverability.4) The resolution of radar images is limited by the inherent objection ofmatched filter and classical time-frequency uncertainty principle.Fractional fourier transform (FRFT), with the good performance of time-frequencyconcentration and compressive sensing (CS), with advantages in incoherence measurement andreconstruction may solve the problem above. Thus, fast ISAR imaging of maneuvering targetsbased on FRFT is firstly discussed in this paper, and then CS and its application in ISAR ismainly studied. The main content is as following:1. Fast ISAR imaging of maneuvering targets is studied, provide that the Doppler variationof scatter sub-echoes can be approximated as a first order polynomial. A novel method namedFast ISAR imaging of maneuvering targets based on Fractional Fourier Transform is suggested.In this new method, FRFT, instead of traditional FFT, is employed to accomplish the highresolution of cross-range. The best Fractional is confirmed following that the Doppler variationhas been quickly estimated by the Ambiguity Function slice of scatter sub-echoes, and thusFractional searching can be avoided and computational efficiency can be greatly improved. Thesimulation results show that high quality ISAR images can be achieved and complexity can begreatly reduced.2. In this paper, a new Modified Sparsity Adaptive Matching Pursuit (MSAMP) Algorithmis proposed for signal reconstruction without prior information of the sparsity. Firstly, a newsparsity estimation method based on atom matching test is used to get an initial estimation ofsparsity. Then it realized the close approach of signal sparse step by step under the frame ofSparsity Adaptive Matching Pursuit (SAMP). But the step size in MSAMP algorithm is variablerather than the fixed one in SAMP algorithm. At the beginning of step iterations, high value ofstep size, causing fast convergence of the algorithm is used initially to realise the coarseapproach of signal sparse, and in the later step iterations smaller value of step size, advancing theperformance of the algorithm is used to achieve the precise approach of signal sparse. Finally, itrealized the precise reconstruction of sparse signal. The analytical theory and simulation resultsshow that significant reconstruction performance improvement is achieved. The problem of overor under estimation in SAMP algorithm under the condition of large sparsity is almost resolved. Also, the convergence of the algorithm is much faster than the fixed step size algorithm.3. CS ISAR imaging algorithm based on sparse representation of range is studied forreducing drang of radar signal processors, resulting from requirements of high resolution andreal-time processing. Firstly, sparse and phase of signals, after stretch processing is analysed, andthen straight random sampling to echos is realized after a phase-reserve sparse representationmodel of ISAR echos under stretch processing model has been established. The analytical theoryand simulation results show that the proposed method can greatly reduce the data rate of highresolution radar imaging system, and the stringent requirement of high speed A/D is almostresolved. Also, the imaging quality is as better as the RD algorithm.4. In allusion to the phenomenon of focus-out and issue of the weakness performance ofCompressive Sensing ISAR imaging based on Fourier dictionary under equably accelerativerotating model, a new method, named compressive sensing ISAR imaging of equablyaccelerative rotating targets is proposed via analyzing echo signals in field of slow time. Firstly,the Chirp dictionary based on matching Fourier transform is constructed after that the ratio ofangular accelerated velocity to initial angular velocity was estimated by some compensatedrange cells. Then, a coherent projection in cross-range is used to improve the SNR ofmeasurements. Finally, CS Recovery Algorithms is used to reconstruct the information ofscatters, and high quality ISAR imaging will be generated after that. The simulation results showthat the proposed method can improve the imaging quality of ISAR significantly and have agood performance of depressing noise.