An improved algorithm for PRI modulation recognition

In modern electronic warfare, it is a severe challenge for the Electronic Support Measure (ESM) system to realize the recognition of the threat radiation source, the modulation of the pulse repetition interval (PRI) is closely related to the function and application of the radar radiation source. Therefore, the recognition of PRI modulation is greatly significance to the identification of the threat radiation source. In order to recognize the PRI modulation type accurately in the presence of pulse missing and noise pulses, an improved recognition algorithm based on neural network is proposed in this paper. In this paper, the features describing a particular PRI modulation type are extracted from the pulse interval, which includes local stationarity, monotonic directionality and symmetry. After 3 dimensional feature spaces are set up, a multi-layer neural network is employed to recognize different PRI modulation type. The simulation results show that the proposed improved method can not only improves reliability and accuracy of recognition, but also greatly reduce the computational complexity.