The Research Of Sorting Method For Multimode Radar Signal

Radar signal sorting is the primary technology of modern electronic countermeasure and future information war, which also is bottleneck technique of the electronic intelligence reconnaissance system and electronic support measure. Radar signal recognition can be implemented only if accurate signal sorting is guaranteed, which is the premise of threat level judgment and appropriate counterplot determination. With the rapid development of modern radar technology and drastic countermeasure of electronic warfare and information war,electromagnetic environment radar signal sorting method faced is becoming increasingly complex and dense, advanced system radar with a variety of work modes, adapted low probability of intercept technology, in-pulse waveform modulation technology, and multi-parameter agility is replacing traditional radar with simple signal form and secular changed or stationary parameters. Aiming at the key existing theoretical problems of signal processing of advanced system radar with multi-mode in complex intensive electromagnetic environment, exploratory research is implemented in this paper, research achievements are as follows:(1) In line with systematic analysis on signal feature of multi-mode radar signals,discussion on the problem existed when traditional sorting model dealing with multi-mode radar signals, and a novel sorting model in allusion to unknown multi-mode radar signals in complex intensive electromagnetic environment is presented to utilize intercepted radar pulse train more effectively.(2) Data mining theory is introduced to radar signal sorting procedure, proposed that describe the multi-mode radar signal with data field theory to barrage the strong interfere of noise and outlier data in complex intensive electromagnetic environment, according to the distribution of equipotential line in data field, nested structure of radar signals can be constructed, which is available for the following hierarchical clustering sorting method. The presented method can avoid parameter setting link in the radar signal sorting procedure.(3) Hierarchical cluster algorithm is introduced to sort multi-mode radar signals,according to hierarchical architecture, hierarchical cluster can form a nested class pedigree chart or class hierarchy by merging or splitting signal sample recursively. Use this property,hierarchical cluster algorithm can conclude signals of different working mode of the same radar emitter to a pedigree, and can effectively solve increasing-batch problem in multi-mode radar signal sorting processing.(4) Use the advantage that cloud model theory can extract qualitative concept in vague,random,uncertain small sample data,proposed a new effectiveness evaluating method for sorting result. In this method,each hierarchical clustering result is regarded as a cloud model,then compare and process the similarity of different cloud model in terms of proposed determinant rule. The method can solve the problem that separating degree of cluster algorithm is not good enough for multi-mode radar signal sorting, and also can feedback and optimize multi-parameter cluster sorting procedure.