【作者】 王旭；

【导师】 保铮； 刘宏伟；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2014， 博士

【摘要】 多输入多输出(MIMO)雷达具有多个发射和接收天线,各发射天线可以发射不同信号,这使得MIMO雷达具有波形分集的优势。根据天线的间距,通常将MIMO雷达分为分布式MIMO雷达和集中式MIMO雷达。在复杂多变的工作环境下,传统雷达所采用的往复扫描的单个窄波束和相对固定的发射波形,很难取得满意的性能,也无法满足多任务、多功能的需求。而MIMO雷达的波形分集优势,使得雷达系统具有更高的发射自由度,可以针对工作场景和需求,灵活的设计发射方向图;可以根据环境信息及时调整发射波形。因此,MIMO雷达方向图和波形设计成为一个很有意义的研究领域。本文从发射方向图逼近、合成信号距离旁瓣抑制、多功能波形、杂波抑制和抗干扰等方面对MIMO雷达波形设计进行了研究。论文的主要工作概括如下:1.从正交波形和具有特定方向图的波形等两方面对MIMO雷达方向图设计进行研究。首先介绍了一种基于梯度下降的正交波形循环设计方法,该方法基于最小化加权积分旁瓣(Weighted Integrated Sidelobe Level,WISL)准则,根据各波形的梯度信息,依次对每个波形进行优化。仿真实验表明该算法与基于最小化峰值旁瓣准则的序列二次规划法性能相当,且可以应用于大规模阵列。针对多波束的方向图设计,利用基于修正的最小化旁瓣模型,设计具有较低旁瓣的多波束方向图;通过调整发射样本顺序和优化初相,有效地抑制了波束方向信号的自相关旁瓣和互相关。最后提出一种基于空时联合优化的MIMO雷达波形设计方法,对集中式MIMO雷达的正交波形设计与具有特定方向图的波形设计建立了统一的优化模型,该方法在实现方向图逼近的同时,可以对感兴趣方向信号的时域自相关旁瓣进行有效抑制。2.初步研究、分析了通过MIMO雷达波形设计来实现同时多功能。首先提出一种MIMO雷达多模波形设计方法,以方向图逼近、波束方向信号的功率谱逼近及频谱逼近等为准则优化波形,所设计的波形可以在期望方向形成多个波束,各波束方向的信号可以具有不同特性,以同时实现搜索、成像和跟踪等功能。在多目标场景中,传统雷达通过波束切换的方式,依次对各目标成像,可减少对于每个目标的驻留时间,从而导致多普勒分辨力的降低。针对此问题,提出一种宽带MIMO雷达波形设计方法,该方法可以实现期望的多波束方向图,波束方向信号具有期望的功率谱特性。通过利用CZT(Chirp-Z Transform)变换来计算各方向信号的离散频谱,有效地降低了计算量。另外,为了在保证分辨率不损失的情况下降低目标回波间的相互干扰,将各期望功率谱随机分布于多个不重叠的子带区域。基于所设计波形,通过应用稀疏重构方法可以实现多目标成像。3.实际场景中目标信息不断变化,需要对波束指向、各目标分配的功率及时调整。针对此问题,提出了一种基于迭代FFT的MIMO雷达快速波形设计方法。该方法分两个独立的阶段:快速方向图综合和空域合成信号优化。首先利用迭代FFT方式设计波形以逼近期望的方向图;然后通过优化波形的初相序列来提高空域合成信号的自相关特性。在初相优化中,采用了最小化积分旁瓣电平(ISL)和加权积分旁瓣电平(WISL)准则来抑制旁瓣。由于该算法的主要计算可由FFT/IFFT实现,因此可以满足实时性要求。4.针对空域合成信号的距离旁瓣抑制,提出了两种方式:脉冲串编码和失配滤波器设计。对于静止或速度较低的目标,脉冲串编码的方式可以在保证方向图不变的情况下,进一步抑制距离旁瓣。而对于较高速的目标,采用一种基于辅助变量的MIMO雷达失配滤波器,可以在一定的信噪比损失(Signal-to-Noise Ratio Loss,SNRLoss)范围下,最小化零频附近多普勒失配下的距离旁瓣。该方式可以在多普勒维展宽距离旁瓣的抑制范围,对多普勒频移具有一定稳健性。5.针对抗干扰和杂波抑制问题,初步研究了MIMO雷达认知发射。为了降低被干扰源侦察到的可能,提出了一种自适应发射置零方法,利用接收到的干扰数据,构造干扰子空间,通过对波形与干扰子空间的正交化,实现自适应发射置零。另外,针对快速运动干扰,通过添加导数约束或利用锥化矩阵可以实现干扰子空间扩展,从而获得零陷展宽。针对空域非均匀杂波的抑制,提出了一种MIMO雷达发射方向图优化设计方法。首先通过发射正交波形对杂波环境进行感知,以发射波形与正交波形回波间的平均相关性作为接收阵列中杂波功率的近似,再根据最大化信杂噪比模型优化发射波形。仿真结果表明,在空域非均匀强杂波背景下,优化后的波形可以有效地提高回波信杂噪比。更多还原

【Abstract】 Multiple-Input Multiple-Output(MIMO) radar consists of multiple transmit antennas and multiple receive antennas. Each transmit antenna can freely choose the transmit waveform, which provide MIMO radar with the advantage of waveform diversity. In general, MIMO radar systems can be classified into two categories according to the configuration of antennas: MIMO radar with widely separated antennas and MIMO radar with colocated antennas. The traditional radar transmits invariable waveform and scans a single narrow beam in the interesting space. In the complex and dynamic environment, the working mode of the traditional radar cannot obtain the satisfactory performance, and cannot meet the requirement for multi-function and multi-task. The waveform diversity supplies more degrees of freedom to MIMO radar. As a result, MIMO radar can freely design the transmit beampattern and adjust the transmit waveform according the operating environment and the requirement. Therefore, MIMO radar transmit beampattern and waveform design become a meaningful research field. This dissertation focuses on the waveform design for MIMO radar, including transmit beampattern design, range sidelobe suppression of the synthesized signals, waveform design with multi-function, clutter suppression and interference rejection. The main content of this dissertation is summarized as follows.1. Transmit beampattern design for MIMO radar is studied in two aspects, i.e. the orthogonal waveform and the waveform with specific beampattern. Firstly, an orthogonal waveform cyclic design method based on gradient descent is introduced. The optimization criterion is to minimize the weighted integrated sidelobe level. By using the gradient information, each waveform is circularly optimized one by one. The performance of the proposed algorithm is similar to that of the sequential quadratic programming(SQP) based on the criterion of minimizing peak sidelobe level. Furthermore, the proposed algorithm can be applied to large scale array. For the beampatten with multiple beams, the minimum sidelobe beampattern design model is modified to obtain the desired beampatten and the low sidelobes. By changing the permutation of the transmit samples and optimizing the initial phases, the auto-correlation sidelobes and the cross-correlation can be suppressed effectively. Finally, a MIMO radar waveform design method based on joint spatial and temporal optimization is proposed. The optimization model is identical for orthogonal waveform and the waveform with specific beampattern. By using the proposed method, the desiredbeampttern can be realized approximately, and the auto-correlation sidelobes of the signals in the directions of interest can be effectively suppressed.2. MIMO radar waveform design for realizing simultaneous multi-function is preliminarily analyzed and studied. Firstly, a multi-mode waveform design method for MIMO radar is proposed. The optimization criterions include beampattern match, power spectrum match or frequency spectrum match at the beam directions. The optimized waveform can form multiple beams at the desired directions. Furthermore, the signals at the beam directions have different properties to simultaneously realize different functions(i.e. searching, imaging and tracking). In the scenario of multiple targets, the traditional radar adopts the approach of beam-switching for each target imaging. This approach may decrease the illuminating time on each target and lead to the degradation of Doppler resolution. Therefore, a wideband waveform design method for MIMO radar is proposed. This method can realize the desired multiple beams and the desired power spectrums at the beam directions. By utilizing the Chirp-Z transform(CZT) to evaluate the discrete frequency spectrum of the spatial synthesized signal, the computational cost is apparently decreased. To reduce the mutual interferences among the echoes from different targets without the loss of range resolution, the desired power spectrums at the beam directions are distributed over multiple unoverlap bands. Based on the designed waveform, multitarget imaging can be implemented by using sparse reconstruction methods.3. As targets informations change continuously in the actual scenario, the beam directions and the powers allocated to different targets should be adjusted timely. Therefore, a fast waveform design method based on iterative FFT is proposed. The waveform design procedure consists of two steps: the fast beampattern synthesis and the spatial synthesized signal optimization. Firstly, the waveform is designed to meet the desired beampattern based on the concept of iterative FFT. Then the initial phases are optimized to improve the auto-correlation properties of the spatial signals based on the criterions of minimizing the integrated sidelobe level(ISL) and weighted ISL(WISL). The dominated computations of the proposed algorithm can be complemented by FFT/IFFT operation, and thus can be performed in real-time.4. To suppress the range sidelobe of the spatial synthesized signal, two approaches are proposed, i.e. the pulse train coding and the mismatch filter design. For the static or lowspeed target, the approach of pulse train coding can effectively suppress the range sidelobe without changing the beampattern. For the high speed target, a MIMO radar mismatch filter based on the instrumental variable is adopted. Under a certain constraint on SNR loss, this mismatch filter can minimize the range sidelobe with Doppler frequency shift near zero. This approach can widen the suppression scope(in Doppler dimension) for range sidelobe and is robust to Doppler frequency.5. Aiming at interference rejection and clutter suppression, the cognitive transmit for MIMO radar is preliminarily studied. An approach of adaptive transmit nulling is proposed to adaptively null the power toward the interferences. Based on the received data, the interference subspace can be constructed. By adjusting the waveform matrix to be orthogonal to the interference subspace, the beampattern can be adaptively nulled at the interference directions. For the rapidly moving interference, the derivative constraints or the covariance matrix taper approach can be used to extend the interference subspace and then widen the nulls. To suppress the clutter in the non-homogeneous environment, a beampattern design method based on a priori information is proposed. Firstly, an orthogonal waveform is transmitted to acquire the clutter distribution information. Then, the clutter power at the receive array can be approximated by the average correlation between the transmitted waveform and the echo of the orthogonal waveform. Finally, the waveform is optimized based on the model of maximizing the SCNR(Signal-to-Clutter-plus-Noise Ratio). The numerical results show that the optimized waveform can efficiently increase the SCNR of echoes in the spatial non-homogeneous environment of strong clutter.更多还原