Research On ELINT Signal Processing Key Technologies For Multifunction Radar

Electronic intelligence(ELINT) reconnaissance plays an important role in morden warefare system. It provides a major way to sense the situation of the battle field and to collect the military and technical intelligence. As radars evolve from fixed analog systems to digitally programmable variants, multifunction radar(MFR) that uses phased array antenna obtains more extensive applications and deployments, and poses a great threat to the targets in the battle field. MFR has the ability to perform multiple functions simultaneously, and operates in an unpredictable manner with a high level of flexibilities, which makes it difficult for the tranditional ELINT system to analyze and decipher the signal of MFR. This dissertation carries out the researches focusing on the key problems brought by MFR in the ELINT system. The content of the dissertation is as follows.In chapter 2, the problem of MFR signal sorting is investigated. Signal sorting is one of the key techniques of ELINT system all along. The signal of MFR must be sorted out from the interleaved signals before the further processing can be implemented. To deal with the problem caused by the agile waveforms of MFR, the signal sorting strategy that uses the information of time difference of arrival(TDOA) is studied. Firstly, the principle of TDOA sorting is elucidated. Then, a model reflecting the distribution of TDOA is built. Based on this, the TDOA distribution features of radars with different PRI type are analyzed quantitively, and the defects of tranditional TDOA sorting method are pointed out as well. Finally, for remedying the defects, a recursive extended histogram based algorithm is proposed. Simulation results show that the proposed algorithm can enhance the performance of TDOA sorting effectively.In chapter 3, the problem of MFR signal modeling and processing is investigated. Firstly, to deal with the problem that the model used in traditional ELINT processing can not cope with the MFR signal, a detailed study of the the signal generation mechanism of MFR is conducted. Based on this, a hierarchical MFR signal model of ―function/task/waveform‖ is proposed. The new hierarchical model can reflect the signal generation mechanism of MFR in a deeper insight. Secondly, a new methodology of ELINT processing of MFR signal is proposed. From the viewpoint of system, an analogy is made between MFR and the organism’s cell, which suggests their high level parallelism in system architecture and policy of operation. Based on this, the methodology and techniques used in the field of bioinformatics, which could be applied for MFR signal analysis, are introduced to construct a new processing framework for MFR signal analysis. The corresponding research methods and mathematical tools are also brought into the domain of ELINT signal processing, which lays the foundation for the further researches.In chapter 4, the problem of MFR waveform unit extraction is investigated. Waveform unit is the basic component of the signal of MFR. So extracting the waveform unit from the received signal is the fundamental and key step of MFR signal processing. Firstly, the detailed study of the signal generation and changing mechanism of MFR is conducted, which indicates that the feature of step amplitude of MFR pulse train caused by the discrete scan pattern of phased array antenna could be used to distinguish the contiguous waveform units one from another. In view of this, the feature of the pulse amplitude is analyzed for both the case of beam dewelling and the case of beam changing. Then, a multiple statistical change point model of the pulse amplitude train is constructed to describe the measured pulse amplitude with noise. Finally, the method of waveform unit extraction is proposed based on the circular binary algorithm, which has been used for abnormal gene detection. The proposed method need no prior knowledges, and can give considerable accuracy of waveform unit extraction even under low signal-to-noise(SNR) condition.In chapter 5, the problem of MFR search regulation reconstruction is investigated. Search regulation reconstruction is the process that separates the signal relevant to search function from the other signals, and find out the rules that regulats the generation of the search signal. Firstly, the signal generation mechanism of MFR is studied to find out the features that can distinguish the signals of different functions, and these features are described in the signal model in waveform unit level. Based on this, the principle is proposed that reconstructs the search regulation by finding the consensus from different periods of the interleaved operational waveform sequence. Then, by introducing the dotplot analysis techniques and multiple sequence alignment(MSA) which have been used extensively in the domain of biological sequence processing, the algorithm of search regulation reconstruction based on sequence similiarity anslysis is proposed. The proposed algorithm can calculate the similairities between different parts of the waveform sequence quantitively, and guarantee a high level of performance for the reconstruction of the MFR search regulation.