【作者】 周升辉；

【导师】 邓维波；

【作者基本信息】 哈尔滨工业大学 ， 信息与通信工程， 2014， 博士

【摘要】 高频地波雷达是一种有效探测海面超视距隐身目标的雷达系统，通常采用宽波束照射目标、单方向窄波束接收回波。由于高频段目标特性处于瑞利区前段或者谐振区，不确定的RCS随机起伏极易导致回波能量微弱而丢失目标。为了克服目标的闪烁，本研究充分利用不同频率和角度目标RCS差异，将多输入多输出（MIMO）体制与高频雷达相结合，利用MIMO体制的优势来提高雷达系统时间、空间、频率管理的自由度，通过有效综合多方向、多载频的回波信息，达到改善系统分辨力的目的，进而提高雷达系统的探测能力。因此，MIMO体制与高频地波雷达结合的关键技术研究具有重要意义。本研究将MIMO的体制结合高频雷达的特点，建立了区别于一般MIMO雷达的多载频MIMO高频雷达模型，分别讨论了集中式和分布式MIMO雷达两种模型，针对不同模型提出了有效的信号处理算法，得到了一些具有积极意义和参考价值的方法和结论。总体来说，本研究主要包括以下几个方面：第一，本文改进了MIMO模糊函数，突破了传统模糊函数理论的局限性，使其适应于复杂的MIMO雷达系统速度距离分辨力的分析评估。集中MIMO模型采用分集发射非连续频点相位编码信号，针对系统空域与时域耦合问题进行了研究，通过扩展的模糊函数有效评估该模型下角度距离分辨力，并得到了非连续频谱发射信号参数准则；进一步在多载频分布式MIMO模型下，通过扩展传统模糊函数的研究评估系统空间分辨力，得到了分布式MIMO雷达的传感器布放规律。模糊函数的扩展改进研究对于MIMO高频雷达全局分辨力分析具有重要的意义，为后续关键问题的提出和研究打下了坚实的基础。第二，为了有效提高高频雷达距离分辨力，针对非连续谱信号产生的较高距离副瓣，本研究提出了两种方法来解决非连续谱信号较高旁瓣问题——接收端非连续谱旁瓣抑制和发射端波形设计。区别于以往非连续谱时分复用构造形式，本研究采用不同天线分集发射不同载频编码信号，接收端将非连续谱信号相参拼接，对相参处理后主瓣附近产生的较高旁瓣进行抑制。基于凸优化算法优化旁瓣抑制滤波器，对有限长数据产生的伪峰进行滑窗剔除，较好的解决了多目标环境下弱目标副瓣遮挡问题。本算法实时性好且简单实用，便于工程实现。基于发射端波形设计，本研究利用波形平均通带阻带功率比与自相关旁瓣关系，建立波形优化的代价函数。通过量子遗传算法优化低自相关旁瓣波形，仿真实验取得了预期的效果，优化的波形实现了频谱约束下最优副瓣等级。本章通过非连续谱信号的设计处理充分利用非连续的频谱资源，改善了MIMO高频雷达恶劣电磁环境的较差分辨力。第三，针对高频雷达的角度分辨力差问题，区别于以往MIMO模型虚拟阵列技术，本研究提出多载频聚焦算法虚拟大的阵列孔径。针对虚拟孔径的单元间距大于半波长，阵列综合产生栅瓣的问题，本研究进一步提出了预条件共轭梯度法重构非均匀空域信息，有效抑制阵列稀疏产生的栅瓣问题。在重构标准频率下均匀空域信息时，需要求解高维方程，预条件的引入显著提高了算法的收敛速度。另外，针对不同频率下，目标不一致的幅度相位响应、回波信噪比及多普勒信息，提出了相应的预处理流程。最后，分析了发射频率的选择对空域综合性能的影响。多载频MIMO空域综合提出了一种新的信号融合思路，有效扩展了高频雷达的孔径尺寸。第四，本研究建立了多载频分布式MIMO高频雷达模型，针对该模型产生空间和多普勒域稀疏采样问题，基于压缩感知提出一种分布式雷达系统算法。本系统采用非均匀周期发射多载频相位编码信号，有效减少多脉冲雷达系统不同节点数据传输负担。针对稀疏回波信号恢复问题，构造统一的空域和多普勒域感知矩阵，有效解决了分布式系统多站点回波信息集中处理问题，及分布式系统产生的目标定位模糊。针对空间网格细化导致压缩矩阵相关性变强，将多载频信息的引入，较好的解决了感知矩阵RIP（空间等距）条件的限制。此外，针对实际情况中，不同方向产生的未知相位响应无法准确重构问题，提出测量矩阵附加目标未知的相位响应来改善目标定位误差。最后，讨论了感知矩阵的列向量相关性，系统估计误差性能与脉冲个数的关系，以及空间分辨网格的划分准则。本研究对于感知压缩算法应用于实际的分布式MIMO雷达模型提供了重要理论依据。更多还原

【Abstract】 HFSWR is an effective surveillance radar system, which can detect the horizonstealth targets over the sea, usually adopts wide beam irradiation, receives the echonarrow beam in single direction. Since in HF band the target characteristics in theRayleigh region or resonance regions, uncertain target attitude and RCS fluctuationscan easily lead to weak echoes and missed targets. To overcome the targetscintillation, the study takes full advantage of RCS differences in differentfrequencies and angles, combining multiple-input multiple-output (MIMO) systemand HF radar, exploiting the advantages of the MIMO system to enhance the radarsystem time, space, frequency management DOF, synthesizing multi-directional,multi-frequency target echo information effectively, to improve the resolution of theobjective system, and improving the system’s resolution. Therefore, the researchabout key technologies of MIMO system and HFSWR combination has importantsignificance.In this study, we combined the MIMO system with HFSWR characteristics, andestablished a multi-carrier MIMO HF radar model which is different from thegeneral MIMO radar. This research established a centralized and distributed MIMOradar model, for different model proposed an effective signal processing algorithms,and obtained some positive significance and reference value methods andconclusions. Overall, this study includes the following aspects:First, the research improved MIMO ambiguity function to adapt complexMIMO radar system speed range resolution analysis, breaking the traditional theoryof ambiguity function limitations. This paper presents a discontinuous spectrumMIMO HF radar ambiguity function, and discusses spatial and temporal couplingunder collocated MIMO model which adopts different carrier frequency diversitytransmission signals, effectively assesses the range resolution of the system, andobtains discontinuous spectrum waveform parameter selection criteria; furtherextends the traditional ambiguity functions to evaluate system spatial resolution, andgets distributed MIMO radar sensor deployment general law in a multi-carrierMIMO model. This study proposes improves MIMO radar ambiguity functionglobally for high resolution analysis, which has important significance for thesubsequent key issues raised and resolved.Second, in order to effectively improve HF radar range resolution, fordiscontinuous spectrum signal generating higher sidelobes shortcomings, this studyproposes two kind of discontinuous spectrum signals to solve higher sidelobe problem-the receiving sidelobe suppression and transmitter discontinuous spectrumwaveform design. Different from the previous time multiplexing structurediscontinuous spectrum waveform, this study adopts antenna diversity to transmitdifferent carrier frequency phase coded signal, the receiver suppress highersidelobes around the mainlobe after signal coherent processing. Based on convexoptimization algorithm we design sidelobe suppression filter to remove pseudo peakwhich is produced by a finite length data sliding window, effectively solving weaktargets are submerged by higher sidelobes under multi-target environment. Theproposed algorithm has good real-time and simple characters, is easy to practicalimplement. Based on the transmitter waveform design, the study obtained therelationship of waveform between average passband-stopband power ratio and theautocorrelation sidelobe level, and proposed waveform optimized cost function. Byquantum genetic algorithm we optimize the low autocorrelation sidelobes waveformto obtain the desired results, the optimized waveform under constraint spectrumachieved the optimal sidelobe levels. This chapter on discontinuous spectrum signaldesign improves HFSWR working performance under harsh electromagneticenvironment.Third, in the perspective of HF radar poor angle resolution, different from theprevious model of virtual MIMO array technology, the study proposed amulti-frequency focus algorithm to synthesize virtual array aperture. For virtualaperture array element interval greater than half the wavelength, the beamformingwill generate grating lobes, this study further proposed based preconditionedconjugate gradient method to reconstruct non-uniform spatial information efficientlyto suppress higher grating lobes. In the reconstruction of the standard frequencyuniform spatial information, high-dimensional equation need to solve, theintroduction of the pre-conditions significantly improves the convergence speed.Meanwhile, for different amplitude and phase response, Doppler and SNR of thedifferent frequencies, we proposed corresponding preprocessesing, and analysis howthe transmitter carrier frequency selection impacts on the overall performance of thespatial synthesize in addition. Multi-carrier MIMO spatial synthesis proposes a newsignal fusion ideas, effectively extends the HFSWR aperture size.Fourth, this study established a multi-carrier Distributed MIMO HF radarmodel, which exists space and Doppler domain sparse sampling problem, andproposed distributed radar system algorithms based on compressed sensing. In thisstudy, we adopt inhomogeneous period to radiates phase-coded multi-carrier signal,that can reduce multi-pulse system radar different nodes’ data transfer burdeneffectively. For sparse echo signal recovery problems, we construct a spatial andDoppler domain in single CS matrix, the algorithm figure out multistatic distributed system centralize processing and distributed systems target location ambiguityeffectively. For mesh refinement causing CS matrix column stronger correlation,multi-frequency information is introduced, it is better to solve CS matrix RIPconditions. Furthermore, in view of the actual situation, unknown differentdirections response results in reconstructing inaccurately, we propose additionalrandom phase CS matrix to improve target location errors. Finally, the studyanalysis the CS matrix column vectors correlation relationship between theestimation error and the number of pulses, and the spatial resolution of the griddivision criteria. This study is important theoretical basis of CS distributed MIMOradar model applied to practical application.更多还原