Long-Time Coherent Integration Based on Reconstructed Spectrum for Weak Target Detection Using Pulse-Doppler Radar

Due to the fact that the target may not be exactly on the sampling cell, the strategy of collecting the energy of a sampling cell closest to the assumed target location adopted in the conventional long-time coherent integration (LTCI) methods may lead to performance degradation. To address this problem, a LTCI method that uses cells around the target location to reconstruct target signal and simultaneously obtain additional intrapulse accumulation gain is proposed. From each cell within the main lobe coverage centered on the target location, a sample at the target location is collected according to the point spread function. All these samples are added to form the reconstructed spectrum, which is then used as target signal to be integrated among pulses. Benefiting from the accurate sampling and spectrum reconstruction, the performance loss induced by the rounding error can be greatly mitigated, and target energy distributed in multiple adjacent cells can be accumulated to acquire additional signal-to-noise ratio (SNR) gain. The integration gain, computational complexity, and multitarget focusing performance of the presented method are analyzed theoretically. Detailed numerical experiments are conducted to demonstrate the effectiveness and improvement of the presented method on integration ability, input-output SNR capability, and detection ability.