Radar Emitter Signal Feature Analysis Based On Time-Frequency Atom

Deinterleaving and recognition of radar emitter signals is an important symbol to measure the technical level of radar countermeasure equipments in modern electronic countermeasure, and hence, it is an urgent problem to solve. Feature parameter is crucial for the recognition and deinterleaving. The lack of new feature parameters seriously restricts the further enhancement of the technical level of our electronic countermeasure equipments. Both as for the technical level's improvement of electronic countermeasure equipments and the radar emitter signals' deinterleaving and recognition, it is of significance to explore effective feature parameters. Time-frequency atom algorithms can extract key information from signals, which have complex or special structures, and capture the nature features of the signals.Consequently, using time-frequency algorithms, the analysis and research have been done to radar emitter signal features. The main work and research fruits are as follows:1. Recent researches and developments of radar emitter signal feature analysis are summarized. And then following the description of mathematical models, the waveforms of radar emitter signals in time-domain are plotted. According to classical time-frequency analysis methods, the properties and characteristics of radar emitter signals are studied.2. The time-frequency atom decomposition is redescribed and two atom dictionaries are provided.3. The greedy algorithms and quantum genetic algorithms are introduced in detail. The experiments are carried out on common radar emitter signals, using the time-frequency atoms built above. According to the experiments, the extracted feature atoms are analyzed. The performances are compared between the different atoms.4. In order to increase the search capability, particle swarm optimization (PSO) is introduced. The experiments carried out on radar emitter signals show that PSO has stronger search capability and better performances than the greedy algorithms.5. According to the above research, a new atom is built, and the corresponding discretization method is presented. Using the atom and PSO, the features of radar emitter signals are extracted. Experiments show that the improved time-frequency atom has good performances.Through the study mentioned above, time-frequency atom features are extracted. At the same time, the experiment results provide the following conclusions.The time-frequency atom algorithms can capture the features of radar emitter signals effectively. The introduction of PSO reduces the searching time of the most suitable atom. The improved time-frequency atom has good performances because the atom can capture radar emitter signal features quickly and validly.