Digital Channelized And Correlative Technology Based On Fourier Transform

With the development of signal processing technology and micro-electronics, Electronic War (EW) receiver has been changed from the analog system to the digital system. The digital channelized receiver based on high speed, large capacity FPGA and DSP technology gets great attention. With the advantages of wide instantaneous frequency-band, large dynamic range and multi-signal processing ability, the receiver can realize full interception of radar signals in monitored frequency band. However there are two bottlenecks in the development of the digital channelized receivers: One is that the speed of signal processing can not keep up with the high sample rate of the ADC devices. ADC sample rate is improved from MHz level to GHz level, which can almost meet the requirement of sampling directly in radio frequency. But the signal processing devices only operate in MHz level, it seriously restricts the processing ability of EW receivers. So the difference between sampling rate and the speed of signal processing is huge. The other is that the frequency resolution of digital channelized receivers is restricted by channel numbers, limited signal duration and process time, which leads to the poor precision of frequency measurement. In order to solve these problems, we must explore new signal processing system.Several new types of channelized receivers based on Fourier transform theory have been investigated in the dissertation, including data-frame sliding STFT algorithm, point-by-point sliding DFT/DTFT algorithm, and controllable recursive algorithm. After deeply study on the Fourier transform algorithm,a channelized receiver with parallel and pipeline processing is constructed. The receiver based on FPGA devices can realize radar signal detection, measurement of radio frequency and pulse parameters, such as signal arriving time, terminative time and pulse width, in real time. These channelized receiver based on Fourier transform have great significance for the miniaturization of smart EW receiver.The main contents and innovations of this dissertation can be summarized as follows.1. The multi-path FFT arithmetic that based on polyphase filter and short FFT technology was presented. The FFT algorithm can be parallel executed by disassembling large point FFT into small point FFT, and then the multi-path FFT arithmetic was applied to receiver based on data-frame sliding STFT algorithm.2. Instead of traditional FFT algorithm, the multi-path parallel recursive algorithm is proposed. In this algorithm, DFT spectrum, DTFT spectrum or ZT spectrum of the finite length data are computed. Several realizable configurations of the recursive algorithm were discussed, including single-phase monopole recursive unit, polyphase monopole recursive unit, single-phase double pole recursive unit, polyphase double pole recursive unit, etc.3. To solve the output instability, the frequency unit of the sliding DFT/DTFT algorithm is reconstructed by the single-phase monopole recursive unit that used to calculate DFT/DTFT spectrum. And then a new type of digital channelized receiver with stable output is constructed using the improved sliding DFT frequency unit.4. The concept of controllable recursive algorithm is proposed. Two kinds of filter banks are made up of the recursive units that calculating DTFT spectrum. And then, they are used in digital channelized receiver to perform real-time interception on radar signal, estimation arrivel time and terminative time of signals, measurement of pulse width and pulse carrier frequency, and so on.5. Using FFT spectrum and spectrum subdivision technology, a synthesized algorithm that combining Rife algorithm and complementary Rife algorithm is proposed. An improved Rife (I-Rife) algorithm is presented by moving the signal frequency to the midpoint area of two neighboring discrete frequencies and then the frequency is estimated by Rife algorithm. I-Rife algorithm not only has high estimation precision, but also has a high stability in all frequency domains. A direct frequency estimation algorithm using DTFT spectrum is discussed. With the amplitudes of multi-spectrum line of DTFT,the universal expressions of the frequency estimation is also given.