Based On Data Association With Multi-receiver Radar Signal Sorting Algorithm

In electronic countermeasure environment, signals are dense and complex, which are new challenges to radar signal sorting. For the purpose of leading the countermeasures system choose the main radiation source and determining the interference means promptly and accurately, pulse sequence derived from the same source of the radiation must be sorted from a series of staggered pulses exactly. The multi-receiver data association can effectively improve the performance of the system, which is of great importance to signal sorting.Pretreatment based on data association, the first main research content is researched. Its purpose is to remove interference noise. Firstly,an improved KNN Algorithm is presented,then the data on two receivers are associated, which is simulated with 100 times Monte Carlo simulation, and the success rate and the error rate are gotted. A new fusion algorithm, D-value comparison on turn is put forward. The data of similar pulses come from same radar is fused by the PDWs , which acquired by 2 receiver sets. Then the original data is recovered. At last the new fusion algorithm is simulated with MATLAB, which prove to be feasible by calculating error rate.Pre-sorting based on clustering algorithm, the second main research content is researched.The results of signal sorting are classified. It includes K-means Algorithm, FCM Algorithm,and BP neural network Algorithm, which are studied in sequence and simulated with MATLAB. Then the application conditions, the advantages and disadvantages of each algorithm are analyzed.Signal sorting, the third main research content is researched, included PRI transform algorithm, the improved PRI transform algorithm, CDIF algorithm and SDIF algorithm. They are studied in sequence and simulated with MATLAB by different parameters. Then the application conditions, the advantages and disadvantages of each algorithm are analyzed. At last, the data with interference noise on two receivers are associated by the improved KNN Algorithm contrast with the same data that have not been associated. So it comes to the conclusion. In the situation of interference, the data that has been associated is sorted, and the correct association is of high probability.