High Resolution Sar Imaging Techniques Under Squint Mode

Synthetic Aperture Radar (SAR) utilizes relative motion between radar sensor and the target being detected to synthesize a wide radar aperture, therefore a fine resolution in azimuth is yielded and so as to image the target.SAR has multiple implementation forms. SAR sensor can be carried on various platforms, such as airplane, satellite and missile. Scan mechanism of SAR antenna can be mechanical or sensor array. The classic working modes of SAR is stripmap, spotlight, scan and the newly developed hybrid mode. Pointing direction of SAR is developed from side looking to squint looking, the later may provide more flexibilities for SAR applications. Nowadays, more than one SAR systems cooperate and distributed in space. Such a multi-SAR system is called constellation SAR and always spaceborne.The relative motion between SAR platform and target will result in Doppler Effect of received signal in azimuth. By accumulating the azimuth phase history, there will be a bandwidth in azimuth. The signals from a same target can be adjusted to a same range cell and the signals can be filtered by a matched filter or some other pulse compression algorithms to yield a comparative azimuth resolution that of range.With the extending of SAR application, it is required that SAR system can generate satisfactory image result under multi modes, with a high resolution, and robust in rigid environment. All of these requirements bring new challenges to SAR signal processing, and shown in following aspects.First, the algorithms are designed for highly squinted SAR data processing from a large swath. As there is a large squint angle between wave beam pointing and broadside, this makes a tight coupling of range and azimuth in range-Doppler domain, therefore a more precise description of SAR signal should be employed for imaging processing. Besides the coupling is range-dependent, a precise model of signal may make the difference of signals between various range cells much more evidence and the depth of focus may be degraded. In current algorithms, a high order (above 3) polynomial model is adopted to approximate phase term of SAR data or the system function is deduced to implement an anti-filtering processing to the data.Second, ultra high resolution algorithms are designed. The range resolution of SAR relates with transmitted bandwidth, therefore it is easy to increase the signal bandwidth to form a fine resolution. Azimuth resolution is decided by Doppler bandwidth. To yield a fine resolution in azimuth, several methods can be taken, such as a shorter wave length, a large beam angle to increase the aperture time, decrease the height of platform from target or increase velocity. Commonly, it is easy to lengthen the aperture time to realize a fine resolution in azimuth. However, an even large range-migration would be introduced.Third, the instability of airborne platform and the trajectory curvature of spaceborne platform. In the ultra high resolution SAR systems, the errors between the actual platform motion trajectory and the ideal one may share a same quantity with a resolution cell. There are high precise sensors of motion in modern flight system, such as pose, position and velocity. It is a key problem of how to utilize those informations in processing SAR data.Last, estimation of parameters, identify and image of motion target. When the high precise motion data of platform cannot be gained, it is the only way to estimate the parameters from the data gathered. The SAR system, which is carried on some simple flight platform such as unmanned aero-plane and small satellite on low orbit, can only be processed through this method. With the increase of SAR resolution, the algorithm to indicate and image motion target on the ground is very important. The classic algorithms is developed according to the Doppler parameters of the stationary target on the ground, therefore, the image of moving target will be unmatched and defocused.This paper will focus on the aspects mentioned above, summarize the current algorithms, study on the problems in highly squinted SAR data image processing. In this paper, following noval algirthms are proposed:Quartic phase based algorithm (QPA). The algorithm finds a balance between the order of polynomial approximating to the coupling of range and azimuth and the performance of dealing with a large squint angle and a wide focus depth. QPA overcomes the shortcomings of existing algorithms, such as a little squint angle endurance and a small focus depth. The complexity of computation is acceptable.Range walk removal based algorithm. The algorithm changes the way to decouple range and azimuth only in Doppler domain, but carry out the decoupling in both azimuth time domain and Dopler domain so as to decrease the complexity of algorithm.Sub-band combination based spotlight SAR imaging algorithm. The algorithm utilizes strip mode SAR processing methods to image each sub-apeture, and then combine azimuth bands of sub-apertures to a full band therefore, recover the full resultion in azimuth.Image quality based algorithms for range-walk estimation and auto-combination of sub-bandsTime-frequency relationship based algorithm for complex oscillation signalAt the same time, the paper also studies on the composition of SAR imaging system and parallel computation techniques. Together with the algorithms listed above, this paper presents a total solution for the squinted SAR image system.