Research On Long-time Integration Algorithms For High Speed Weak Targets

The high speed maneuvering low observable targets in modern war havechallenged modern radars’ detection ability. Long-time integration technique is aneffective method to raise the SNR of radar echo and improve the radar’s detectionperformance. Under the gaze work mode of PD Radar, this paper focused on theinvestigation on three respects of problems: the echo model construction ofrectilinearly moving targets with uniform velocity and acceleration, the analysis ofechoes’ across range units effect (ARU) and Doppler frequency migration (DFM)effect, the investigation of long-time coherent integration algorithms. By worksabove, we intend to enhance the performance of echo integration and improve theradar’s capability of detecting high speed weak targets without knowing their motionparameters.Under the signal model of rectilinearly moving targets with uniform velocity,long-time coherent integration algorithm based on Keystone transform (KT) andRadon-Fourier transform (RFT) was introduced respectively. A Doppler ambiguityfactor estimation algorithm based on Radon transform is proposed, whoseeffectiveness was verified through numerical experiments. The numericalexperiments reached conclusions below: Keystone transform can complete thecompensation of envelop walk blindly. However, it would become tedious whendealing with multi-targets. Radon-Fourier transform can realize the extraction oftarget’s track and integration of target echo energy by jointly searching along rangeand velocity directions of moving targets. RFT can avoid the problem of Dopplerambiguity factor estimation. But the problem of blind speed side-lobe needs to benoticed.Under the signal model of rectilinearly moving targets with uniform acceleration,long-time integration algorithms based on Keystone transform with second-orderphase compensation, joint second-order Keystone transform and RFT, and thesecond-order generalized RFT was proposed. As to the problem of Doppler frequencymodulation rate estimation, algorithm based on Dechirping method, Modified Chirp-Fourier transform and Fractional Fourier transform was introduced. Numericalexperiments verified the effectiveness of these methods. The numerical experimentsreached conclusions below: algorithms based on KT can’t avoid the problem ofDoppler ambiguity factor estimation. So they are limited when dealing with multi-targets. Algorithm based on joint second-order KT and RFT can applied in theprocessing of multi-targets. But the estimation of Doppler modulation rate requireshigher SNR. The second-order generalized RFT can coherently integrate the energy of moving targets by jointly searching the range, velocity and acceleration of thetarget. But the burden of computation makes it hard to implement in project.