Research On Detection And Parameter Estimation Of LFMCW Signals For Radar Reconnaissance

Linear frequency modulated continuous wave(LFMCW) radar is widely used in some fields, such as military navigation, battlefield reconnaissance, imaging etc., due to its simple architecture, its small volume, its high range resolution, its none range blind, its low cost, its low energy transmitted and its low probability of intercept. As a result of the continuous wave system, LFMCW radar has a lower transmission power compared to the pulse Doppler radar, which leads to the lower signal to noise ratio(SNR). In order to ensure the effectiveness of signal detection, more sampling points are required. Therefore, high computational complexity and low SNR become the major factors of constraining the effective detection and parameter estimation of LFMCW signals. The key of signal detection and parameter estimation is to reduce the computational complexity and improve the SNR. From the perspective of signal processing, some topics related to signal detection and estimation are comprehensively analyzed and elaborated in this dissertation, such as the signal model, parameters estimation and detection for LFMCW signals. The main work is summarized as follows:1. The method for detection and parameter estimation of mono-component periodical LFMCW signal is researched. Firstly, two kinds of transmitted models and intercept models are introduced. Secondly, for the problem of high computational complexity and low SNR, a method based on the joint window-variant coherent average method, the cyclostationary method and the dechirp processing is proposed to detect mono-component periodical LFMCW signal and estimate its parameters. Based on the periodicity of signal, the improvement of the SNR and the estimation of the modulation period and the chirp rate are realized by the window-variant coherent average method and the cyclostationary method at first; then, the initial time-offset, the initial frequency, the intial pahse and the amplitude of signal are extracted with the dechirp processing, based on the estimated modulated period and chirp rate. This method realizes long time accumulation with low computational complexity and improves the SNR greatly. However, the performance of detection is limited by the range of searching window. For the problem above, an improved method is proposed. The improved method modifies the winow-variant coherent average method, then, uses the improved winow-variant coherent average method and cyclostationary method to estimate the modulation period and the chirp rate, later, utilizes the dechirp processing to estimate the initial time-offset, the initial frequency, the intial phase and the amplitude. The detection performance of the improved method is no longer restricted by the range of searching window. Furthermore, the improved method can also realize long time accumulation with low computational complexity, which greatly improves the SNR. Simulations verify the effectiveness of the proposed method.2. The method for detection and parameter estimation of multi-component periodical LFMCW signal is researched. Considering quick realization of detection and parameter estimation on low SNR condition, a method based on the inter-frame method, the cyclostationary method and the dechirp processing is proposed to detect multi-component periodical LFMCW signal and estimate its parameters. Firstly, an inter-frame correlation method and its equivalent form are successively proposed based on the periodicity of signal, which can achieve a linear superposition of multi-component singal energy by simple complex additions and complex multiplications. Secondly, according to the characteristic of the inter-frame correlation, a target peeling method is introduced, which can achieve the detection of signal and estimation of the modulated period. Finally, based on the idea of the CLEAN algorithm and the estimated modulated period, the chirp rate, the initial time-offset, the initial frequency, the initial phase and the amplitude of each component are estimated by the cyclostationary method and the dechirp prosessing. The method realizes long time accumulation with low computational complexity, which greatly improves the SNR, and achieves the detection of each component without two-dimensional search, So the method has lower calculation. Simulations verify the effectiveness of the proposed method.3. For the problem of quick realization of detection and parameter estimation on low SNR condition, a method based on the zero frequency searching method, the cyclostationary method and the dechirp processing is proposed to detect LFMCW signal and estimate its parameters. Firstly, a zero frequency searching method is introduced based on the instantaneous correlation property of LFMCW signal, which can realize the signal detetion and eatimation of the modulated period. Secondly, an improved mehod is introduced, which grealy decreases the amount of calculations of zero frequency searching method while maintaining the same performance. Thirdly, based on the estimated modulated period, the chirp rate is estimated by the cyclostationary method. Finally, the dechirp processing is utilized to estimate the initial time-offset, the initial frequency, the initial phase and the amplitude, based on the estimated chirp rate and modulated period. The proposed method realizes long time accumulation with low computational complexity, which greatly improves the SNR. What’s more, the method can realize the detection and estimation of the muti-component LFMCW signal under certain conditions. Simulations verify the effectiveness of the proposed method.4. For the problem of high computational complexity and low SNR, a method for detection and parameter estimation of multi-component LFMCW signal is proposed, based on the matched inter-frame correlation method, the cyclostationary method and the dechirp processing. Firstly, a matched inter-frame correlation method is introduced based on the inherent characteristic of inter-period, which can realize the signal detetion and eatimation of the modulated period according to the difference of the modulated period and the initial frequency between each component. Then, based on the idea of the CLEAN algorithm and the estimated modulated period, the chirp rate, the initial time-offset, the initial frequency, the initial phase and the amplitude of each component are estimated by the cyclostationary method and the dechirp prosessing. All components can be detected and estimated respectively according to the amplitude. The proposed method not only effectively realizes the detection and parameter estimation of LFMCW signal on low SNR condition, but also has low computational complexity in favor of engineering implementation. Simulations verify the effectiveness of the proposed method.