Real-time PRF selection for medium PRF airborne pulsed-doppler radars in tracking phase

This paper proposes a new method to select optimal pulse repetition frequency (PRF) sets for use in tracking mode of medium PRF airborne pulsed-Doppler radar. Neural networks algorithm is used to map from engagement variables to the optimal PRF set. On-line computation during flight can be made real-time after off-line training of the neural network. The training sets for the neural network need to be generated by selecting optimal PRF set for the possible engagement scenarios from which range-Doppler clutter map is calculated to check the decodability and detectability for all PRF candidates. The PRF sets generated by the method must guarantee the maximum detectability inside the target tracking window as well as maintaining good decodability. Simulation result shows that the proposed method has much better range-Doppler detection performance compared to the previous algorithms by applying different optimal PRF set to different engagement scenarios and target positions.