Bernoulli filter with signal coherent integration for target detection and tracking in clutter environments

In practical applications, the target maneuvers might lead to high-order range migration (RM) and Doppler frequency migration (DFM) effects, which seriously degrade the performance of long-time coherent integration. In this case, the target detection and tracking performance will deteriorate due to insufficient signal energy and heavy clutter. To address this problem, a recursive algorithm based on the Bernoulli filter with long-time coherent integration is proposed in this paper for maneuvering target detection and tracking in low signal-to-noise ratio environments. The Keystone transform is used to correct the linear RM. Then, the matched filtering is performed based on the state to eliminate the high-order RM and DFM. According to the focused peaks in the range-time azimuth-Doppler domain, the target existence probability and dynamic states are estimated using an amplitude-aided Gaussian mixture Bernoulli filter. Simulations are performed to demonstrate the effectiveness of the proposed algorithm. Compared with traditional coherent integration methods based on parameter searching, the proposed algorithm acquires close integration performance with less computational complexity, and the target detection and tracking performance is improved with the aid of integrated signal amplitude.