Study On Long-time Coherent Integration For Radar Maneuvering Targets

Increasing the integration time is an effective method to improve maneuvering targetdetection performance in radar applications. However, range migration and Doppler spreadcaused by maneuvering target motion during the integration time make it difficult to im-prove the coherent accumulation of target’s energy and detection performance. This disser-tation studies new techniques for range migration correction and Doppler spread compensa-tion and parameter estimation of maneuvering target during the long integration time.The main content of this dissertation is summarized as follows．1. A new parameter estimation method named MDP-KTRFT is proposed to solve therange migration and Doppler spread caused by the high-speed maneuvering targetmotion in the mono-target scenario. This method utilizes the second-order Keys-tone transform to correct the range curvature, and the improved dechirping methodto compensate for the Doppler spread. Radon Fourier Transform (RFT) is thenused to correct the range walk for target coherent detection and high-precision pa-rameter estimations of the target can be obtaind.2. A new parameter estimation method named FrFT-KTRFT is proposed to solve therange migration and Doppler spread caused by the high-speed maneuvering targetmotion in the multi-target scenario. This method utilizes the second-order Keys-tone transform to correct the range curvature, and the Fractional Fourier Transform(FrFT) to compensate for the Doppler spread. RFT is then used to concentrate theenergy of targets and estimate the parameters of targets. MDP-KTRFT andFrFT-KTRFT can correct the range migration and the time-varying Doppler fre-quency of the target without knowing the motion information of targets. Also, thetwo methods can effectively avoid the influence of Doppler ambiguity on integra-tion so that it can improve the precision of parameter estimation of high-speedmaneuvering target sharply. Compared with the second-order RFT algorithm, thecomputational burden of the proposed methods is greatly reduced under the pre-mise that they have similar estimation accuracy of range, velocity and acceleration.3. A new joint estimation algorithm for multi-targets’ motion parameters based onKeystone transform and Lv’s Transform (LVT) is proposed, which can imple-mented without using any searching operation. This method can resolve the prob- lem that the precision of parameter estimation of conventional LVT algorithmwould be affected by the range migration. The signal-to-noise ratio threshold forthe proposed method is also given.4. Since the parameters of high-speed target might be out of the range of LVT esti-mator, an improved method based on dual-carrier frequency processing and LVT isproposed to ensure the LVT estimator works well for high-speed target. In thismethod, two sub-band signals are designed with different center frequencies andthe shift and conjugate processing are implemented to obtain the composite signal.After that, the Keystone transform and LVT processing are applied on the compo-site signal. The simulation and experimental processing results demonstrate thatthe new method can estimate parameter precisely without any searching operationand can correct the range walk of multi-targets simultaneously, resulting in the lowcomputational burden.5. A new parameter estimation method based on segmental Keystone transform andLVT in the frequency domain is proposed to estimate parameters of ground mov-ing targets without velocity ambiguity. In this method, the segmental Keystonetransform is first applied to correct the range walk of targets and FFT is then ap-plied within each signal segment in the azimuth dimension. After that the samefrequency resolution bins of each segment are selected to construct new series andKeystone transform is used to correct the frequency walk. Finally, inter-segmentLVT is implemented to obtain the parameter estimates. The Keystone transformand LVT processing can be implemented through parallel processing, making thenew method suitable to be applied in memory-limited and real-time processingsystems. Simulation results demonstrate that the proposed method can estimate theparameters of targets witout a priori knowledge on the target motion parameters.6. For the ground moving targets with velocity ambiguity, a new parameter estima-tion method based on ambiguity number searching is proposed. This method firstemploys the segmental Keystone transform to correct the range walk of targets andthen estimates the velocity ambiguity number using the image entropy criterion ordirectly searching processing. After that, FFT is applied within each signal seg-ment in the azimuth dimension and the same frequency resolution bins of eachsegment are selected to construct new series. Keystone transform is then used to correct the frequency walk. Finally, inter-segment LVT is implemented to obtainthe parameter estimates. This method can obtain the parameter estimates of targetsonly with the searching of the velocity ambiguity number, which is verified by si-mulated and real data.