Study On SAR/ISAR Focusing Approach For Complicated Motion Condition

Synthetic apreture radar and inverse synthetic aperture radar(SAR/ISAR)have the ablitiy of long time observing under the condition of all-weather,day and night.They can extract the topographic information and physical dimension of the observation scenarios and targets by digital signal processing methods effectively.SAR and ISAR methods have been widely applied in the national defense techniques and civil industry,such as topographic mapping,battlefield reconnaissance,disaster forecast,space stuation awarness and air defense antimissile system,etc.For SAR/ISAR imaging technique,high 2D resolution of range-azimuth is the key to representing the detailed characteristics of the targets.The range resolution depends on the transmitted signal,and the azimuth resolution depends on the length of synthetic aperture.For the existing SAR system,two-dimension resolution is limited by the radar system and difficult to be improved.Combining with the platforms which have samll size,such as unmanned aerial vehicles and helicopters,SAR can select the observing scenarios in more flexible way by utilizing scanning or spotlight modes,the resolution of images and the sheltering can be improved by increasing the observing angles.But the platforms always move with complex trajectory to meet the requirement of multi-angle observing,which brings challenges to the subsenquent imaging processing,and these platforms with small size are very sensitive to the weather and airflow,we need to study the efficient and robust algorithms to compensate the motion errors.Moreover,the high sidelobes and ring effect appear in the images,which obtained by long observing time and wide rotating angle,is another problem need to be solved.With the development of minitype aerial vehicles,it has significant vlaues to study on real-time ISAR imaging algorithms with high resolution for these maneuvering targets.To solve the key problems and technological difficulties of complicated motion condition in SAR/ISAR focusing,we expand the work from real-time imaging,high resolution and motion error compensation methods for ROSAR,CSAR and non-uniformly rotating targets.The relevant work is supported by the National Basic Research Program of China(973 Program)under Grant 2011CB707000 and the National Natural Science Foundation of China under Grant 61231017 and 91438106.The main contributions and works of this dissertation are summarized as follows:1.Under the condition of high resolution,range cell migration induces the complex expression of slant range history and two-dimensional wavenumber spectrum,which brings the challlenges in subsequent imaging processing.To solve these problems,we proposed the modified wavenumber domain imaging method for ROSAR,it can correct the range migration precisely and obtain the well-focused images under the condition of wide beam and large scenario.Due to the traditional ROSAR algorithm which expands the range history to second order cannot meet the requirement of high resolution,a novel ROSAR imaging method based on Cardano's formula is proposed,which corrects the RCM terms and obtain the well-focused ROSAR images.2.Because ROSAR is usually applied in the platforms with small size,such as unmanned rotorcraft or helicopter,it is very sensitive to the alteration of airflow and vibration of platform,which may induce serious motion error and deteriorate the imaging perfermance significantly.To deal wih this problem,an autofocusing algorithm based on ROSAR wavenumber imaging is proposed.We first design the expented ?-k method for ROSAR to correct the RCM and make it easier to perform the subsequent motion error compensation.Then we combine the quadratic phase correction with the phase gradient estimation,and eliminate the local linear components via breaking the ROSAR data into small subaperures.After several iterations,the motion errors can be estimated and compensated,hence,the well-focused ROSAR image can be obtained.3.In BCSAR systems,the limited signal bandwidth induces serious ring effect and high sidelobes to solve these problems.To solve these problems,we propose a novel imaging scheme named as GD-BCSAR,and furtherly study its Fourier sampling area and image resoltion.GD-BCSAR can obtain the larger Fourier sampling area,and the spectrum discontinuity of Fourier sampling space can also be reduced,which suppresses the ring effect and high sidelobe in BCSAR images effectively.4.In ISAR imaging for non-uniformly rotating target,the time-varying Doppler frequency induced by the target complex motions will degrade the quality of ISAR image seriously.Considering that most of existing ISAR methods always suffer from the heavy computational burden and considerble propagation errors,which deteriorates the performances of algorithms and limits their practiacal applications.To tackle these problems,an efficient ISAR iamaging algorithm based on geometry-aided parameters estimation for non-uniformly rotating targets is proposed.The 3-dB filtering and weighted least square estimation are utilized to eliminate the interference of noise and cross-terms,hence,the estimation accuracy of derivative of chirp rate can be improved.Similarly,the chirp rate and center frequency can be estimated with high accuracy via WLSRT in TFD domain and range-Doppler domain,respectively.The proposed method can eliminate the effect of propagation error and reduce the computational complexity simultaneously,and achieve the real-time ISAR imaging of non-uniformly rotating target.