Research On Airbone AESA Signal Recognition Based On Electronic Reconnaissance Phased Array

Signal that airborne AESA(Active Electronically Scanned Array) emits has wide bandwidth, low SNR(signal-to-noise ratio), changing characters and so on, which makes it difficult to capture and identify the signal via traditional electronic reconnaissance method. In response to the challenge that the airborne AESA signal’s less capture builds to the electronic reconnaissance capacity of our country, this thesis performed some research on recognition method and scheme designing of AESA signal based on the electronic reconnaissance array.Aimed at the airborne AESA’s properties of wide bandwidth and low SNR, the multiple targets directions finding method of super-resolution in frequency based on the branch band acquired from frequency spectrum’s support domain of electronic reconnaissance array with incoherent accumulation is proposed. First, the FFT of each channel’s digital sampling signal on base band along shutter time transforms the signal to frequency domain. Second, pile up the accumulation of those details in frequency along space domain to realize incoherent accumulation. Third, search for all the peak of the frequency spectrum after accumulation, sort them, acquire the average as a reference threshold using a portion of smaller peaks and achieve the frequency points beyond the threshold as frequency spectrum’s support domain of AESA signal. Then, partition frequency spectrum’s support domain on the basis of some portion of aperture fill width and set multi-channel data in frequency point of branch band as snapshots in frequency. At last, realize the DOA(direction of angle) estimation with super-resolution within each branch band with snapshots in frequency and IAA(iteration adapt algorithm) in incoherent processing mode and obtain the spectrum estimation for wideband of the whole frequency spectrum’s support domain by accumulation. The simulation result based on ULA(uniform line array) with 56 elements presents that the proposed method hold a good capability of DOA estimation with super-resolution when the SNR is below-20 d B, if the wide band signal doesn’t have mixture of frequency spectrum.Faced the problem that it’s hard for AESA signal classification in the background with complex electron and magnet, it’s realized based on DOA estimation of signal by the structure in temporal beam-forming for wide band that the enhancement in spatial of AESA signal and the suppression of other electronic reconnaissance signal. The simulation result indicates that the structure in temporal beam-forming for wideband can separate multiple wideband signals effectively and signal classification is achieved. On the basis of electronic reconnaissance signal classification with beam-forming for wideband, recognition of multi-parameter of the airborne AESA signal, including time-of-arrival(TOA), wide of pulse(PW), pulse repeat interval(PRI), bandwidth of spectrum in frequency(bandwidth) and the mid-frequency(0f) of signal.Empowered by the key technique on the recognition of airborne AESA signal, the implement scheme of hardware about recognition of wideband signal is showed. At first, with the precondition of the recognition of wideband signal, the whole frame of signal processing system structured by FPGA and multi-core DSP processor is planned with each module’s function provided and the procedure of data processing is discussed. Then, the design of state machine for the system is accomplished and the three states of self-check, function and correction presented in detail. At last, in response to the case of the multiple targets directions finding method in frequency with super-resolution based on the branch band got from frequency spectrum’s support domain with incoherent accumulation proposed, the procedure of software for the method with multi-core DSP processor TMS320C6678 is designed and the distribution of multi-core’s mission and communication between cores are discussed.