Research On Probability Association-based Signal Sorting Method

Radar signal sorting has always been a key technology and research hotspot in electronic warfare.The variety of signals,rapid changes in radar parameters and high signal density are three characteristics of complex electromagnetic environment,these will cause high density pulse,similar radar parameters,pulse overlap,clutter and other phenomena.These phenomena make it difficult to sort radar signals.Based on the urgent needs of the radar signal sorting,this essay makes a thorough study of radar signal preprocessing and main processing respectively.The main contents are summarized as follows:Firstly,aiming at the similar or overlapping of different radar parameters,a radar parameter probabilistic clustering model is proposed.It is based on joint probability data association algorithm.The radar parameters are transformed from the measurement domain to the probability domain,so that the deviation between pulse description word and different radar parameters is converted to the value of the associated probability.At the same time,when a pulse description word falls into the parameter's thresholds of several different radars at a sample time,the radar parameters are updated by the association probability,namely,the clustering center of the radar parameters are updated.This algorithm can avoid the influence of clutter on the clustering center and effectively improve the accuracy of the clustering center.Secondly,to improve the radar parameter probabilistic clustering model,a multihypothesis analysis algorithm is proposed.The proposed algorithm is based on the multihypothesis tracking algorithm.The algorithm generates hypothesis on events which associates pulse description word from different radars.Moreover,it makes a reasonable analysis of the pulse overlap situation,uses the characteristics of pulse overlap to generate hypothesis.Use radar parameter probabilistic model to update the parameters of possible radar trees.Then after a hypothesis is generated and the parameters are updated by N times,the algorithm backtracks the authenticity of the hypothesis,confirms the most reasonable hypothesis,and removes other hypotheses.Simulation results show that the algorithm further improves the correct matching rate of clustering.Finally,this paper studies joint sorting algorithm,it joints multi-hypothesis analysis clustering and the improved SDIF algorithm.Step one,improve SDIF algorithm,and enhance its ability to deal with jitter radar signals,clutter and interference.Then test the radar signal multiple hypothesis analysis and improved SDIF joint sorting algorithm.The simulation results indicate that it has a good correct sorting rate in situations that contains the mingling of radar parameters,pulse overlap,and electromagnetic interference with clutter.