Research On Highly Squinted Imaging Algorithms For Helicopter-borne Rotating Syntnetic Aperture Radar And Bistatic Synthetic Aperture Radar

With the ability of day&night and all weather condition imaging on observing areas,synthetic aperture radar (SAR) plays an important role in earth observation,environment monitoring, military reconnaissance and safety fighting and landing foraircrafts. To improve its system capability and extend its application platform,researches on the novel SAR systems have been carried out in the SAR community.Azimuth synthetic aperture can be obtained based on the the radar antenna rotating.Thus, helicopter-borne rotating synthetic aperture radar (ROSAR) can achieve twodimensional imaging. It has more advantages such as high spatial and temporalresolution, short revisit period, forward-looking imaging capacity, simple structure andlow cost, and two dimensional images can be obtained without the necessity of anymovement of the helicopter platform. The observation scene’s three dimensional imagescan be acquired by combining the forward movement of the helicopter platform and therange pulse compression technique. Bistatic SAR is a kind of synthetic aperture radarwhose transmitter and receiver are mounted on difference platforms. Compared withmonostatic SAR, bistatic SAR can obtain the targets’ non-backscattering features,making the design of the transmitter and receiver more flexible and beneficial for thereceiver platform’s stealth design. Specifically, highly squinted bistatic SAR canobserve the targets of.certain areas without flying through it. That is, highly squintedbistatic SAR has the capability of imaging on the front target. Therefore, as a newimaging configureation for the front targets, helicopter-borne ROSAR and bistatic SARcould be widely used in both military and civilian fields.Sponsored by the National Natural Science Foundation of China and China’s BasicScientific Research Project (973), this dissertation researches the principles and imagingalgorithms of the helicopter-borne ROSAR and bistatic SAR. The key innovations inthis thesis are as follows:1. Considering the dependence of the azimuth resolution of the helicopter-borneROSAR on the azimuth reconstruction angle, and trational subaperture processingazimuth resolution suffers from low resolution problem, a novel echo signal model isestablished using a higher order range model by analyzing in detail the helicopter-borneROSAR geometric configuration. Then the accurate two-dimensional spectrum isderived by series reversion. Based on which, a new high-resolution imaging algorithmis proposed for Helicopter-borne ROSAR, detailed imaging procedure and the expression of the compensation factor are also presented. Moreover, the impacts ofsystem parameters selection on azimuth resolution are also analyzed.2. An improved Chirp Scaling algorithm incorporating motion compensation forhelicopter-borne ROSAR is proposed to accommodate the helicopter platformmovement during radar antenna rotating imaging. Firstly, the geometric model ofhelicopter-borne ROSAR with platform movement is established. Based on which, therelationship of the slant range error introduced by platform movement variation with thetarget position within the synthetic aperture time is analyzed in detail. Then, the slantrange error’s range spatial characteristics can be obtained and correspondingcompensation method is given. Next, the impact of the range-dependent velocity causedby the helicopter-borne ROSAR inherent geometric configuration on imaging quality isanalyzed and the compensation methods are incorporated into the range cell migrationspatial variance correction and azimuth compression. With only FFT and complexmultiplication and no interpolations, the proposed method can be efficientlyimplemented. Finally, simulation results demonstrate that well-focused SAR image canbe obtained for wide swath scenarios and existing platform movements using theproposed method.3. A novel helicopter-borne ROSAR imaging model and algorithm based onfrequency modulated continue wave (FMCW) are proposed. Firstly, using the principleof equivalent phase center, the separated transmitting and receiving system is turned tobe a monostatic one. Based on which, precise two-dimensional spectrum is deduced,besides Doppler offset caused by the continuous motion of the antenna during thetransmitting and receiving is analyzed and compensated. Then efficient inverse Chirp-Ztransform scaling is applied to correct the range-dependent range cell migration, besidesthe effects of range-dependent velocity approximation errors on the imaging algorithmare analyzed and corresponding compensation method is also given. With only FFT andcomplex multiplication and no interpolations, the proposed method can be efficientlyimplemented. Finally, correctness of the analysis and effectiveness of the proposedalgorithm are demonstrated through simulation results.4. Combining the helicopter platform movement and the radar antenna mounted onat the tip of the rotator blades rotating movement, the helicopter-borne ROSAR canachieve three dimensional images for the platform’s front objects. In this thesis, a novelforward-looking three-dimensional imaging algorithm is proposed for helicopter-borneROSAR based on the special imaging geometric configuration and the characteristics ofthe echo signal. Firstly, the azimuth and range well-focused SAR image relative to the same area can be obtained by using the improved chirp scaling algorithm, and thenevery along-track and slant-range slice data is regarded as a collection from thedownward-looking imaging mode. Based on that, an improved omega-k algorithmincorporating subscene processing technique is developed to obtain accurate along-trackand slant-range focusing images. Then, the three-dimensional SAR image can beobtained by processing all the slices with the same procedure. The validation of theproposed method is done by exploiting simulated data. Since one only antenna is usedto get the targets’ three-dimensional image, the helicopter-borne ROSAR has low costand simple structure characteristics.5. Apart from the range shift caused by the linear range walk correction, for thehighly squinted azimuth-variant bistatic SAR, another severe range offset is introducedby the inherent azimuth-variant geometric configuration and the impacts becomesignificant with the increase of the azimuth resolution and the difference between thetwo unparalleled tracks. On the other hand, the azimuth–dependent higher orderquadratic FM rate terms and cubic phase term become important with the increase of thesquint angle. To accommodate for this problem, an extended azimuth nonlinear chirpscaling algorithm (EANLCSA) is investigated in this thesis. Firstly, range-azimuthcoupling is mitigated through a linear range walk correction operation, and then bulksecondary range compression (BSRC) is implemented to compensate the residual rangecell migration and cross coupling terms. Following which, the characteristics of theazimuth-dependent quadratic and cubic phase terms are analyzed, and modified scalingcoefficients are derived by adopting higher order approximation and incorporating theazimuth-dependent range offset caused by the inherent bistatic geometric configuration.Compared with traditional nonlinear chirp scaling method, large azimuth depth offocusing can be realized without changing the overall procedure.