【作者】 陈莹；

【导师】 池灏；

【作者基本信息】 浙江大学 ， 物理电子学， 2015， 博士

【摘要】 宽带信号的获取成了现代雷达,射电天文学,电子对抗以及另外一些先进的通信系统必不可少的技术。传统基于电器件的信号获取具有带宽小,功率损耗大以及易受电磁干扰等缺点。光子学技术由于具有带宽大、重量轻、损耗低以及抗电磁干扰等诸多优点,成为未来宽带信号获取的一个重要发展方向。目前,基于光子学宽带信号获取的两种重要技术是光子模数转换技术和光子压缩感知技术。本论文中,我们以提高信号获取的性能为目的,包括精度更高,带宽更大,提出多种基于光子模数转换和光子压缩感知的宽带信号获取方案。论文首先介绍了微波光子学的研究背景、主要应用方向和重要器件。然后针对光子学技术在宽带信号获取中的应用进行了研究,介绍了基于光子模数转换和光子压缩感知的宽带信号获取以及与相应的电技术相比存在的优势。针对现有技术存在的问题,提出了一些创新的结构和方案,通过仿真和实验对系统进行了验证。论文的主要创新点和学术贡献如下：1.通过分析光子模数转换经典模型(Taylor方案)存在的限制,提出了一种基于差分编码的光子模数转换方案,并通过数值仿真对系统进行了验证。方案中利用了一个相位调制器(PM)和一个干涉仪并结合波分复用技术实现了输入信号差分形式的编码,克服Taylor方案中调制器电极长度成倍增长的问题,从而提高了系统的可行性。另外,差分编码的实现提高了系统的等效量化等级。为了进一步提高系统的性能,我们利用双驱动的非平衡调制器代替PM和干涉仪,不仅可以减少系统的体积,降低系统的损耗,也有利用系统的集成。2.在深入研究对称数字系统(symmetrical number system, SNS)的编码特性以及SNS在模数转换中的应用的基础上,提出了改进的基于SNS的光子模数转换方案。方案中利用等半波电压的干涉仪以及多个电比较器,并通过设计干涉仪之间的相移,实现了具有格雷码性质的SNS编码,从而大大提高了系统的纠错能力。为了消除光源幅度抖动的影响,提出将平衡检测技术应用到系统中,从而提高了系统的稳定性。3.在充分理解压缩感知基本原理的基础上,用理论研究和数值仿真的方法详细分析了基于光子压缩感知的宽带信号获取。介绍了基于空间光调制器(SLM)和电光调制器(EOM)的光子压缩感知方案,通过实验验证了系统的可行性以及存在的问题。提出基于平衡马赫增德尔调制器(MZM)结构的随机调制方式,简化了压缩感知系统的数学模型,实现了零均值的随机观测矩阵,提高了系统的性能。4.针对压缩感知系统中,随机序列的速率需达到输入信号带宽两倍的限制,提出基于光时域拉伸技术的压缩感知方案。光时域拉伸技术充分利用色散导致的群延时对加载在啁啾光信号上的射频信号进行拉伸处理。将其与压缩感知技术相结合不仅可以降低输入信号的速率,从而降低系统对随机序列速率的要求,而且可以进一步降低系统的采样速率。为了消除光强度调制和直接检测引入的直流,进一步提出基于平衡结构的光时域拉伸子系统,提高了信号获取的信噪比。5.提出三种基于微波光子滤波的压缩感知方案,在光域中实现压缩感知中的低通滤波或累加过程。方案一利用多波长的非相干光源和色散介质,在群速度色散的作用下实现了随机调制信号相邻比特位之间的累加。方案二利用单波长的连续光源和色散介质,实现了随机调制信号的低通滤波,简化了系统。方案三利用集成芯片的技术实现了基于光频率梳妆的多抽头微波光子滤波器,增加了滤波响应的旁瓣抑制比,减少了信号恢复的误差,有利于系统向集成化的方向发展。通过实验和数值仿真验证了系统实现宽带稀疏信号获取的性能。更多还原

【Abstract】 Wideband signal acquisition has become an essential technique in modern radar, radio astronomy, electronic warfare (EW) and other advanced communication systems. Conventional electronic solutions are limited with narrow bandwidth, high power consumption and susceptible to electromagnetic interference (EMI). To solve these problems, photonic techniques are preferable candidates due to its various advantages, such as high bandwidth, light weight, low loss, and immunity to EMI. Currently, photonic analog-to-digital conversion (ADC) and photonic compressive sensing (CS) are two major approaches of photonics-based wideband signal acquisition. This thesis is targeted at improving the performance of signal acquisition, including higher resolution and wider bandwidth. Several schemes of wideband signal acquisition based on the technique of photonic ADC and photonic CS are proposed and demonstrated.In this dissertation, the research background, major applications and important devices of microwave photonics are firstly introduced. Then, photonic techniques in the application of wideband signal acquisition is studied. The advantages, of photonic ADC and photonic CS based wideband signal acquisition compared to their electronic counterparts are analyzed. At last, some innovative structures and techniques are proposed and demonstrated to solve some existed problems. The major innovations and contributions are as follows:1. For overcoming the major limitation of classical model of photonic ADC, which uses electro-optic modulators (EOMs) with geometrically scaled half wave voltages, differentically encoded photonic ADC is proposed and demonstrated. In the scheme, a phase modulator (PM) and an interferometer are applied, which improves the feasibility of the ADC system. In addition, the realization of differential endcoding can increases the equivalent quantization level of the system. In order to improve the performance of the system, we replace the PM and interformeter with a dual-drive unbalanced modulator, which can reduce the size and loss of the system and is also benefical to the integration of the system.2. Based on the study of symmetrical number system (SNS) based encoding property, a novel photonic ADC scheme based on SNS is proposed and demonstrated. In the scheme, interferometer array with identical half wave voltages and several electronic comparators are applied. By setting the phase shift of adjacent interferometers, SNS with Gray code property are realized, which enhance the error correction capability of the system. To eliminate the amplitude jitter of the optical source, the technique of balanced detection is applied to improve the stability of the system.3. Wideband signal acquisition based on photonic CS is studied. The feasibility and existed limitations of spacial light modulator (SLM) and EOM based photonic CS is demonstrated and analyzed. Optical random mixing with balanced Mach-Zehnder modulator (MZM) structure for photonic CS is proposed, which simplify the mathematical model and the realized zero-mean measurement matrix can improve the performance of the system.4. One disadvantage of CS based signal acquistion is the system bandwidth is limited as applied random sequence should not less than twice of the signal bandwidth. To overcome this problem, photonic CS with the technique of photonic time stretch is proposed and demonstrated. To eliminate the dc component introduced by intensity modulating and direct detection based optical link, photonic time stretch with balanced structure is further proposed. System performance under different parameters is analyzed though numerical simulations.5. Three schemes of photonic compressive with microwave photonic filter to realize the process of low pass filtering or integration of CS are proposed and demonstrated. In the first scheme, incoherent multi-wavelength optical source and dispersive medium are applied and the integration of random mixed signal is realized with effect of group velocity dispersion. In the second scheme, a continuous-wave and dispersive medium is used to simplify the structure and the random mixed signal is low pass filtered with the function of dispersion induced RF power fading. In the third scheme, a multi-tap microwave photonic filter is realized by utilization of an optical frequency comb and a dispersive medium. The optical frequency comb is achieved by spectrally shaping of an amplified spontaneous emission (ASE) source with a silicon micro-ring resonator. The realized filter function has higher side-lobe suppression and better frequency response due to the controllable comb number. In addition, the use of compact on-chip component is beneficial to the integration of the system.更多还原