【作者】 张勋；

【导师】 柯亨玉； 万显荣；

【作者基本信息】 武汉大学 ， 无线电物理， 2021， 博士

【摘要】 外源雷达是一种自身不发射电磁信号,而是利用第三方非合作辐射源进行目标探测的双/多基地雷达系统,具有隐蔽性好、抗干扰能力强、成本可控等诸多优点。无线通信技术的发展为外源雷达提供了丰富的辐射源,其中数字广播信号以带宽稳定、近似“图钉”的模糊函数等特点,成为外源雷达的理想辐射源之一。基于数字广播信号的外源雷达是近年来新体制雷达的研究热点。外源雷达一般设有两个通道,一个参考通道用于接收参考信号,另一个监测通道用于接收目标回波信号。因为发射信号未知,一般通过对重构后的参考信号与杂波抑制后的监测信号进行互相关模糊函数(匹配滤波)计算,以获取目标的距离多普勒信息。参考信号质量的好坏直接影响时域杂波抑制和匹配滤波的效果,从而影响外源雷达探测性能。如何获取使得探测性能最优的重构参考信号是外源雷达信号处理的基础。数字地面多媒体广播(DTMB)是我国拥有自主知识产权的数字电视地面广播传输标准。目前该信号在全国大多数城市已经实现了覆盖,这为我国数字广播外源雷达(Digital Broadcasting-based Passive Radar,DBPR)的研究提供了良好的条件。DTMB信号采用了时域同步正交频分复用(Time Domain Synchronous Orthogonal Frequency Division Multiplexing,TDS-OFDM)调制技术,具有传输效率高、抗多径干扰能力强等优点。基于数字调制的特征,DTMB信号一般可通过“解调-再调制”的方式进行重构。本文针对DBPR参考信号重构中存在的一系列问题,展开了深入研究,包括:时域相关信道估计算法不够稳健且不易并行实现、存在采样频率偏差时参考信号重构、等带宽采样导致高频混叠、如何提纯参考信号使得时域杂波抑制效果最优。具体内容如下:(1)针对DTMB信号中基于伪随机(Pesudo Noise,PN)序列的时域相关信道估计算法不够稳健且不利于并行实现的问题,提出了一种改进的时域相关信道估计算法。该算法先将接收信号与本地帧头信号进行互相关计算,然后利用帧头信号自相关函数的数值特征,获得一组关于信道时域响应的线性方程,最后直接通过线性方程组求解得到信道的时域响应。因为不需要设置门限及迭代干扰消除,改进后的算法更稳健计算效率更高,且不会降低信道估计的精度。仿真和实测数据验证了该方法的有效性。(2)针对存在采样频率偏差时,重构参考信号使得时域杂波抑制效果恶化问题,提出了一种基于整体均衡的参考信号重构方法。首先剖析了造成时域杂波抑制效果下降的原因,即在重构参考信号中帧头使用本地已知信号,帧体使用信道均衡后的信号,导致二者包含不一样的采样频率偏差;然后构建了一种包含循环前缀的数据结构,将帧头和帧体作为一个整体进行信道均衡,从而得到基于整体均衡的重构参考信号,避免了破坏原发射信号结构。相较于传统重构方法,此参考信号重构方法使得时域杂波抑制效果更优,从而能有效提高目标探测性能。最后通过仿真和实测数据验证了该改进算法的有效性。(3)针对等带宽采样下高频混叠问题,提出了一种过采样下的参考信号重构方法。首先分析了等带宽采样的信号特性,及高频混叠对雷达探测的影响;然后在使用过采样导致时域相关信道估计算法失效的前提下,从理论上揭示了过采样下最小二乘(Least Square,LS)算法无法估计信道时域响应的原因,并提出了两种改进的LS算法解决信道时域响应无法被估计的问题,进而获得过采样下的重构参考信号;最后通过仿真和实测数据验证了该方法的有效性。(4)针对重构参考信号与实际发射信号失配的问题,提出了一种基于杂波抑制效果的参考信号重构方法,亦称为可信重构。首先介绍了参考信号重调制原理,并根据工程应用建立了新的信号模型;然后分析了重调制参考信号与监测通道时域杂波抑制效果之间的关系,从理论上证明了存在一个重调制参考信号可使监测信号获得最优的时域杂波抑制效果,并给出了相应重调制参考信号的获取方法;最后仿真和实测数据验证了可信重构方法的有效性。更多还原

【Abstract】 Passive radar is a bistatic/multistatic radar system that does not transmit electromagnetic signals but exploits the non-cooperative illuminators of opportunity(IOs)for target detection.Passive radar possesses the advantages including covert operation,strong anti-interference ability,controllable cost and so on.The development of wireless communication technology provides a wealth of illuminators for passive radar.Among them,digital broadcast signals have the characteristics of stable bandwidth and ambiguity functions similar to "thumb-like",making them one of the ideal illuminator for passive radar.Digital broadcasting based passive radar has become an edging research field of new radar systems in recent years.Passive radar generally has two channels,one is the reference channel used to receive the reference signal,and the other is the surveillance channel used to receive the target echo signal.As the transmitting signal is unknown,the cross-correlation ambiguity function(matched filtering)is performed on the reconstructed reference signal and the clutter-suppressed surveillance signal to obtain the range and Doppler information of the target.The quality of the reference signal directly affects the results of time-domain clutter suppression and matched filtering,thereby affecting the detection performance of passive radar.The reconstruction of the reference signal which brings the best detection performance is the basis of passive radar signal processing.Digital terrestrial multimedia broadcasting(DTMB)is a digital television terrestrial broadcasting transmission standard with independent intellectual copyright in China.At present,the signal has wide domestic coverage which provides good conditions for Digital Broadcasting-based Passive Radar(DBPR)research in China.DTMB signal is a digital TV broadcasting signal using Time Domain Synchronous Orthogonal Frequency Division Multiplexing(TDS-OFDM)modulation technology.It has the advantages of high transmission efficiency and strong anti-multipath interference ability.DTMB signal can generally be reconstructed by “demodulationremodulation” procedure based on the features of digital modulation.In this article,a series of issues in the reconstruction of DBPR reference signal are studied in depth,including the time-domain correlation channel estimation algorithm is not robust enough and not easy to implement in parallel,the reference signal acquisition under non-ideal factors,equal bandwidth sampling causes high-frequency aliasing problem,and the reconstruction of the reference signal that optimize the time-domain clutter suppression.The details are as follows:(1)To make the time-domain correlation channel estimation algorithm based on the pseudo-noise(PN)sequence in the DTMB signal more robust and work in parallel,an improved time-domain correlation algorithm is proposed.The algorithm first calculates the cross-correlation between the received signal and the local frame header signal.Then,the numerical characteristics of the frame header signal autocorrelation function are used to obtain a set of linear equations involved with the channel timedomain response.Finally,the channel time-domain response can be obtained directly by solving the linear equations.As the improved time-domain correlation algorithm avoids the iterative interference cancellation process and does not need to set the threshold value,the robustness and the calculation efficiency of the algorithm are improved.Meanwhile,the accuracy of channel estimation is not reduced.Simulation and field experiment verify the effectiveness of the proposed algorithm.(2)To relieve the deterioration of time-domain clutter suppression brought by the reconstructed reference signal under the sampling frequency offset,a reference signal reconstruction method based on overall equalization is proposed.First,it analyzes the reason that harms the time-domain clutter suppression effect.In the reconstructed reference signal,the frame header uses the local known signal,and the frame body uses the signal after channel equalization,which causes the two contain different sampling frequency deviations.Then,a data structure including the cyclic prefix is constructed.Channel equalization is performed on the frame header and the frame body as a whole,thereby obtaining a reconstructed reference signal based on the overall equalization,avoiding damage to the original transmission signal structure.Compared with the traditional reconstruction method,the improved reference signal reconstruction method can achieve better clutter suppression effect and effectively improve the target detection performance.Finally,the effectiveness of the improved algorithm is verified by simulation and field experiment.(3)To overcome the high frequency aliasing due to equal bandwidth sampling,a reference signal reconstruction method under oversampling is proposed.First,the signal characteristics of equal bandwidth sampling and the influence of high frequency aliasing on radar detection are analyzed.Then,under the premise that the time-domain correlation channel estimation algorithm fails due to the use of oversampling,the reason why the Least Square(LS)algorithm cannot estimate the channel time-domain response under oversampling is theoretically revealed.Two improved LS methods are proposed to solve the problem that the channel time-domain response cannot be estimated,and the reconstructed reference signal under oversampling is obtained.Finally,the effectiveness of the method is verified by simulation and field experiment.(4)To solve the mismatch between the reconstructed reference signal and the actual transmitted signal,a method of reference signal reconstruction based on the clutter suppression effect is proposed,which is also called trusted reconstruction.First,the reference signal re-modulation method is introduced.And a new signal model based on engineering applications is established.Then the relationship between the remodulation reference signal and the time-domain clutter suppression effect of the surveillance channel is analyzed.It is theoretically proved that there is a re-modulated reference signal so that the surveillance signal can obtain the best time-domain clutter suppression effect,and its acquisition method is given.Finally simulation and field experiment verify the effectiveness of the trusted reconstruction method.更多还原