Study On Wideband Interception Methods

With increasing complexity and denseness of battle electromagnetic signals in modern EW, the research of signals reconnaissance,such as signal interception, parameter estimation,and signal sorting is one of the major tasks in electronic reconnaissance under the unkown electromagnetie circumstance. With the development of digital receiver and array signal processing, it is possible to use many processing algorithm flexible to ensure the wideband receiver. A series of key technologies of wideband receiver in complex electromagnetic environment, including algorithm employed to optimize channelized receivers,signal detection and coding, algorithm of non-aliasing parameters estimation with spatial-temporal sub-Nyquist sampling, etc., are mainly researched in this dissertation.This dissertation is devoted to research the algorithm of temporal-frequency -spatial wideband reconnaissance. The main contributions can be summarized as follows.In the aspect of wideband receiver, technology of digital channelization based on STFT (Short Time Fourier Transform) is discussed. The charateristics and existing problems of an efficient channelized receiver structure with weighted overlap-add filterbank to be solved are pointed out. Based on above discussion an architecture based on polyphase decomposition is proposed,which can realize frequency measurement with high accuracy under the condition of the same system resource. In the aspect of all-central frequency of subbands with uniform distribution, a recursive algorithm is proposed. Performance comparison has shown that the proposed structure is of relatively more flexible and higher efficiency. Modern electronic interception systems perform the tasks of detection and main parameters estimation for radio pulse signal. The encoder clusters all the feature vectors belonging to the same signal and estimates PDW (Pulse Descriptor Word).For the problem of frequency aliasing with temporal sub-Nyquist sampling,a method based on cross spectral interpolation from two independent sampling channels for frequency estimation is proposed. Considering spurious peaks when multi signals simultanoues existing, a method of select frequecy range around each cross spectral peak before using MUSIC(Multiple Signal Classification)algorithm to complete frequency search is provided, which can solve the problem of aliasing zone location and improve frequency estimation. Two methods mentioned above are all working on two sub-Nyquist sampling channels. A novel frequency estimation approach based on an auxiliary time-delay channel and ESPRIT(Estimation of Signal Parameters via Rotational Invariance Techniques) algorithm for wideband signal with sub-Nyquist sampling is proposed. With the aid of information provided by the auxiliary delayed sampling channel, we call solve the frequency aliasing. Comparison via simulation has shown that the proposed algorithm is of relatively low computational complexity and the same high performance.The problem of unambiguous angle estimation is studied. The theory of phase comparison approach and the factors that can affect the measurement errors are discussed based on the mathematical model of traditional phase interferometer. An algorithm to estimate DOA(Direction Of Arrival) based on crosspower spectrum phase is proposed and a necessary condition of array antennas configuration is also provided. The two approaches all perform DOA estimation with cross-correlation phase. The estimation will not be precise because of the interferences among multi signals. This problem is solved by using digital channelized receiver. For multiple noncoherent signals, a novel unambiguous DOA estimation is proposed by exploiting the characteristic of space-time DOA matrix.To multi parameters estimation, the information structure of the polarization-phase interferometer is analyzed, two dimension DOA and polarization estimation are presented. The wideband model of the array outputs with spatial-temporal sub-Nyquist sampling is discussed.And then, joint parameters estimation for wideband LFM signal based on delay and auto-correlation calculation is performed.To multi signals interference, TRD (Time-Frequency Rate Distribution) of each sensor can be obtained. Because the TRD of LFM signal possesses the cluster characteristic related to the frequency rate,the spectrum peak in the TRD expresses signal frequency rate.The phase difference of each sensor at the same spectrum peak is the product of the frequency rate and time delay.We have proved that combination TRD transformation with phase comparison approach can be implemented for joint estimation.In conclusion, this dissertation describes several methods of unambiguous parameters estimation in the wideband receiver.All these methods can be treated as wideband interception methods.