【作者】 申东；

【导师】 张林让；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2013， 博士

【摘要】 受多输入多输出(MIMO)通信和综合脉冲孔径雷达(SIAR)的启发，MIMO雷达作为一种新体制雷达，得到了雷达界的广泛关注和研究。MIMO雷达使用多个发射天线分别发射不同的信号，同时使用多个接收天线接收并处理目标回波信号。与传统雷达相比，正是这种波形分集能力使得MIMO雷达具有更多的优势。波形分集是MIMO雷达的重要特征，波形设计是实现波形分集的重要手段。基于混沌信号的新体制雷达是混沌应用于雷达的一个主要方面。代入不同初值，在短时间内能够产生数量众多的混沌信号。混沌信号具有对初始值敏感、类似噪声的宽频谱、图钉型的模糊函数、尖锐的自相关和近似正交的互相关等特性。本论文围绕MIMO雷达混沌波形设计展开相关研究，主要包括以下几个方面：1）提出了混沌离散频率编码、混沌二相编码和混沌四相编码的MIMO雷达波形设计方法以及选取依据。对常用的四种离散混沌序列（Tent序列、Logistic序列、Quadratic序列和Bernoulli序列）的自相关和互相关特性进行了研究，给出了混沌序列的相关性与序列长度、自由参数之间的关系；对使用四种混沌序列调制所产生的离散频率编码波形、二相编码波形和四相编码波形的自相关和互相关特性进行了研究，给出了混沌波形的相关性与波形长度、自由参数之间的关系。以下结论作为选取MIMO雷达混沌波形的依据：Tent和Bernoulli离散频率编码波形的相关性较好，是较为理想的MIMO雷达发射波形；四种混沌二相编码波形的相关性相似，都是较为理想的MIMO雷达发射波形；Tent、Logistic和Quadratic四相编码波形的相关性较好，且相关性好于二相编码波形，是较为理想的MIMO雷达发射波形。2）建立了基于混沌离散频率编码波形和混沌相位编码波形的MIMO雷达多脉冲压缩累积信号处理模型。对混沌波形的自相关旁瓣和互相关的统计功率峰值进行推导：取不同的波形长度N时，自相关旁瓣和互相关的统计功率峰值近似为1N；对不同初始值所产生的混沌波形的自相关旁瓣之间的相关性、互相关之间的相关性进行了研究并得到以下结论：不同初始值所产生的混沌波形的自相关旁瓣之间是不相关的，同时，不同初始值所产生的混沌波形的互相关之间也是不相关的；基于上述相关性结论，建立了基于混沌离散频率编码波形和混沌相位编码波形的MIMO雷达多脉冲压缩累积信号处理模型。使用多脉冲压缩累积能够有效降低混沌波形的自相关旁瓣和互相关。3）提出了用混沌离散频率编码波形提高MIMO雷达目标检测性能的方法。根据混沌离散频率编码波形的相关性与波形长度、自由参数之间的关系，设计并选取较长的且具有较好自相关和互相关特性的Tent离散频率编码波形。在杂波和噪声背景下，脉压输出具有较低的旁瓣峰值，从而提高MIMO雷达对点目标的检测性能，但相同方法却不能提高MIMO雷达对相邻弱小目标的检测性能。通过多脉冲压缩累积的方法，基于Tent离散频率编码波形的MIMO雷达信号处理模型，降低混沌波形的自相关旁瓣和互相关，使输出结果具有较低的旁瓣峰值，从而提高MIMO雷达对点目标的检测性能。特别在相邻弱小目标场景中，由于采用了多脉冲压缩累积的处理方法，强目标回波的旁瓣并不会掩盖弱目标回波的主瓣，从而提高了系统对相邻弱小目标的检测能力。4）提出了用混沌相位编码波形提高MIMO雷达目标检测性能的方法。在杂波和噪声背景下，根据混沌相位编码波形的相关性与波形长度、自由参数之间的关系，设计并选取Tent相位编码波形。由于较长的混沌相位编码波形具有较好的相关性，此时脉压输出具有较低的旁瓣峰值，从而提高MIMO雷达对点目标的检测性能。同时，此方法对相邻弱小目标检测也是有效的。基于混沌相位编码波形的MIMO雷达信号处理模型，使用多脉冲压缩累积降低混沌波形的自相关旁瓣和互相关，从而提高MIMO雷达对点目标和相邻弱小目标的检测性能。更多还原

【Abstract】 Inspired by multiple-input multiple-output (MIMO) communication technique andsynthetic impulse and aperture radar (SIAR), as a new radar, MIMO radar is proposedand becomes the research focus concerned by scholars from many countries. MIMOradar systems transmit different signals via multiple antennas, receive and process targetechoes with multiple receiving antennas. Compared with traditional radar, MIMO radarhas more advantages with waveform diversity. Waveform diversity is an importantfeature of MIMO radar, and waveform design is a mean to achieve the waveformdiversity. New radar systems based on chaotic signal is a main field of radar applyingchaos. With different initial values, a large number of chaotic signals can be produced ina short time. The chaotic signal is sensitive to initial values, with a noise like broadbandspectrum, thumbtacked ambiguity function and sharp auto-correlation andapproximately orthogonal cross-correlation properties.In this dissertation, chaotic waveform design for MIMO radar is investigated, andthe main research focus on the following issues:1) The waveform design and choosing method for MIMO radar are proposed forchaotic discrete frequency coding, chaotic binary phase coding and chaotic four phasecoding. The auto-correlation and cross-correlation properties of four kinds of discretechaotic sequences (Tent sequence, Logistic sequence, Quadratic sequence and Bernoullisequence) are studied and the connections between correlation properties of chaoticsequences and sequence length, free parameter are given. The auto-correlation andcross-correlation properties of discrete frequency coding waveforms, binary phasecoding waveforms and four phase coding waveform from four kinds of chaoticsequences are analyzed and the connections between correlation property of chaoticwaveforms and waveform length, free parameter are given. Chaotic waveform choosingfor MIMO radar accords to the following results, the correlation properties of Tent andBernoulli discrete frequency coding waveforms are good, they are relatively idealtransmitting waveforms for MIMO radar. The correlation properties of four kinds ofchaotic binary phase coding waveforms are the same, all of them are relatively idealtransmitting waveforms for MIMO radar. The correlation properties of Tent, Logisticand Quadratic four phase coding waveforms are good, and better than the binary phasecoding waveforms, they are relatively ideal transmitting waveforms for MIMO radar.2) Signal processing model of multiple pulse compression accumulation for MIMOradar based on chaotic waveforms is established, including chaotic discrete frequency coding waveforms and chaotic phase coding waveforms. Auto-correlation sidelobes andcross-correlation statistical power peak of chaotic waveforms were deduced. Withdifferent waveform length N, auto-correlation sidelobes and cross-correlation statisticalpower peak are approximately1N. From different initial values, the correlationproperties of auto-correlation sidelobes and cross-correlation of chaotic waveforms arestudied, the following results are given: From different initial values, it is uncorrelatedbetween auto-correlation sidelobes of different chaotic waveforms; meanwhile, it isuncorrelated between cross-correlation of different chaotic waveforms too. According tothe results about correlation properties, signal processing model of multiple pulsecompression accumulation for MIMO radar based on chaotic waveforms is established,including chaotic discrete frequency coding waveforms and chaotic phase codingwaveforms. Using multiple pulse compression accumulation can reduce auto-correlationsidelobes and cross-correlation of chaotic waveforms effectively.3) With chaotic discrete frequency coding waveforms, an improvement method onperformance of MIMO radar for target detection is proposed. According to connectionsbetween correlation properties of chaotic discrete frequency coding waveforms andwaveform length, free parameter, longer Tent discrete frequency coding waveforms withgood auto-correlation and cross-correlation are designed and chose. In the clutter andnoise background, output of pulse compression has low peak sidelobe level, and thenthe performance of MIMO radar for point target detection is improved. But, the samemethod can not improve the performance of close small target detection for MIMOradar. Based on the signal processing model for MIMO radar with Tent discretefrequency coding waveforms, multiple pulse compression accumulation can reduceauto-correlation sidelobes and cross-correlation of chaotic waveforms, the output haslow peak sidelobe level, and then the performance of MIMO radar for point targetdetection is improved. Especially in close small target scene, due to multiple pulsecompression accumulation for processing, the sidelobes level from the strong targetecho does not cover up the mainlobe level from the weak target echo, so the detectionperformance of close small target is improved.4) With chaotic phase coding waveforms, an improvement method on performanceof MIMO radar for target detection is proposed. In the clutter and noise background,according to connections between correlation properties of chaotic phase codingwaveforms and waveform length, free parameter, Tent phase coding waveforms aredesigned and chose. Because of good correlation properties for longer Tent phasecoding waveforms, the output of pulse compression has low peak sidelobe level, and then the performance of MIMO radar for point target is improved. This method is validfor close small target detection too.Based on the signal processing model for MIMOradar with chaotic phase coding waveforms, multiple pulse compression accumulationcan reduce auto-correlation sidelobes and cross-correlation of chaotic waveforms, andthen the detection performance of MIMO radar for point target and close small targetare improved.更多还原