Pulse deinterleaving using isomorphic subsequences

The estimation of PRIs (Pulse Repetition Interval) is the core of pulse sorting. Conventional pulse sorting algorithms, based on accumulation (such as FFT), have two shortcomings: one is its difficulty in extracting short pulse segments that are dispersed in massive pulse streams; the other is the heavy computation burden for real-time processing. Accounted for interleaving and missing of multiple pulse groups, the presented algorithm consists of two main parts: repetitive substrings searching and TOA difference matrix processing. Compared with other pulse deinterleaving methods, our algorithm is robust to jitter noise and missing pulses. The computation complexity of the presented algorithm is approximately O(N).