Study On Imaging Algorithms For Circular Trace Scanning SAR

Synthetic aperture radar(SAR) is a microwave remote sensing technique working all day long under all weather to achieve the wide swath and high resolution images of large area earth surface and detect the buried objects. With the development of the SAR technology in recent years, it has been appllied into many practical field, the requirements of SAR performance become much higher due to the growing number of observation missions, such as applications in underground resource detection, battlefield surveillance, and disaster rescue, especially for the large area accurate reconnaissance in a short time, which brings a new challenge for the system design in resolution and swath width.Circular trace scanning SAR(CTSSAR), as a new fast large area SAR imaging mode, maintains not only the advantages of the conventional SAR but also the features of short revisit period and large visual angle imaging, becoming a current research hotspot. In CTSSAR, the flight platform follows a circular flight path at a certain altitude plane with a uniform speed. The antenna is fixed pointing outside the circle all the time. After a whole circular flight, there will be an anular area illuminated by the wavebeam. As the azimuth scanning speed of CTSSAR increases, the illuminated area will be larger than that of the conventional stripmap SAR imaging mode, which contributes to the fast large area imaging. It can be applied to many practical field.However, different from the traditional stripmap SAR mode, the straight line assumption adopted by the conventional SAR full-aperture imaging methods cannot fit the real flight path due to the impact of the circular trajectory of CTSSAR, eventually resulting in the slant range approximation errors so that the final image is defocused. Taking the key techniques and problems of CTSSAR into consideration, this paper mainly focuses on CTSSAR high resolution imaging, squint CTSSAR, frequency modulated continuous wave CTSSAR and so on. In addition, the accurate two-dimensional spectrum of CTSSAR echo signal is derived and the higher order slant range approximated imaging algorithm is designed, realizing precise CTSSAR focusing. The main research achievements are as follows.The first part studies the improved range Doppler imaging algorithm based on the Cardano’s formula. Due to the issue of high order slant range approximated errors caused by the circular trace in full aperture CTSSAR high resolution imaging, the cardano’s formula is adopted in the derivation of the two-dimensional spectrum expression of the echo signal to obtain the stationary phase point. By utilizing the principle of stationary phase, the two-dimensional spectrum for CTSSAR is obtained, an improved range Doppler imaging algorithm is proposed accordingly with the discussion of resolution analysis. In the end, a motion compensation idea of combining inertial navigation system output and autofocus method is introduced.The second part proposes the modified approximated Omega-K CTSSAR imaging algorithm based on series reversion. In order to get a more widely suitable imaging algorithm, the method of series reversion is adopted to obtain the CTSSAR two-dimensional spectrum. The method uses Doppler extended expression to reversely derive the stationary phase series coefficients, based on which the signal’s spectrum can be consequently achieved. This method can control the precision of the spectrum expression by adjusting the order of the series, which contributes to the future extended application. A modified approximated Omega-K imaging approach is designed based on the derived spectrum. The simulation results validate the feasibility and effectiveness of the spectrum and the proposed imaging algorithm.The third part focuses on a modified Chirp Scaling algorithm suitable for large area CTSSAR. In order to realize fast wide swath high resolution imaging, it is necessary to take the range variation of range history in each range unit into account. Considering the motion characteristics of the CTSSAR platform, the high order approximated range equation is employed. The modified Chirp Scaling method is adopted to compensate the difference of range cell migration for each range unit. The bulk range cell migration is then implemented to complete the large area CTSSAR high resolution imaging. There is no interpolation operations included in the approach and thus leading to a high computational efficiency. The point target simulations demonstrate that the proposed algorithm can focus the wide swath CTSSAR image accurately.The fourth part studies the squint CTSSAR mode and its characteristics and proposes an effective imaging algorithm. The squint CTSSAR geometry model is established first, and its special observation is analyzed. Some parameters are discussed in detail and compared with those of squint CTSSAR and classical straight line SAR. Through the comparison and analysis, their inner relationship can be revailed. According to the special model of squint CTSSAR, a modified Range/Doppler squint CTSSAR imaging algorithm based on the modified hyperbolic range equation is proposed. In order to overcome the squint CTSSAR spectrum derivation problem, this method builds a modified hyperbolic equivalent range equation to approximate the real squint CTSSAR range equation, by adding a linear modified factor to the conventional hyperbolic range equation. Moreover, combining with the principle of stationary phase, a squint CTSSAR two-dimensional spectrum is provided, and the corresponding imaging algorithm is also given, completing precise focusing.The fifth part is contributed to the frequency modulated continuous wave CTSSAR(FMCW-CTSSAR) modified approximated Omega-K imaging algorithm based on the residual phase correction. The FMCW-CTSSAR echo signal model is established. Different from the conventional SAR, FMCW-CTSSAR needs to take the residual quardratic phase errors caused by continuous motion and the circular trace based cubic phase errors into consideration. A modified approximated Omega-K FMCW-CTSSAR imaging algorithm based on the residual phase correction is proposed, considering the range errors resulted from circular trajectory. And the image deterioration caused by the residual quardratic phase errors is also analyzed. A corresponding compensation function is provided to remove the impact of the residual quadratic phase errors. The whole algorithm is only composed of FFTs and phase multiplications, which increases the computational efficiency significantly. Point target simulations validate the effectiveness of the proposed imaging algorithm.