【作者】 崔晶；

【导师】 汪立新；

【作者基本信息】 杭州电子科技大学 ， 通信与信息系统， 2012， 硕士

【摘要】 自跳频技术问世以来，以其抗干扰、低截获率、高频谱利用率等优点广泛应用于军事通信领域。随着跳频信号的跳宽越来越大，由于受限于奈奎斯特采样理论，在超高频段通信系统应用中，现有的硬件技术水平在跳频信号的侦察、检测方面遇到了众多困难。近年来出现的压缩采样理论突破了奈奎斯特采样定理的限制。该理论指出，如果信号在某些变换域表现出稀疏性，可以采用少量的、非相关的测量完成对信号的低速高效采集。针对目前在超宽带跳频信号的检测过程中遇到的众多困难，将压缩采样技术应用到跳频信号的采集过程中，研究跳频压缩采样的AIC技术。但是，跳频信号的大宽带必然导致众多定频干扰信号的混入，使得跳频信号的稀疏度大为增加，严重影响跳频压缩采样的效率。由于直接压缩采样法“先重建后抑制”导致硬件和算法上的困难，论文提出了后验压缩采样法、堙灭滤波法和最小均方误差法三种“先抑制后重建”的定频抑制技术。并根据这三种定频抑制方法，构建了一种基于精简测量矩阵的跳频压缩采样结构模型，给出了解决超宽带跳频信号采集的新思路。论文主要内容包括四部分。第一部分介绍了跳频技术的基本原理以及传统检测方法中实现定频抑制的功率谱对消法；第二部分从信号的稀疏表示、测量矩阵的设计和信号的重建这三大问题对压缩采样理论展开论述，并仿真验证了该理论的可行性；第三部分介绍了直接型和预调制型两种主要的AIC实现结构，基于超宽带跳频信号的特点，仿真结果表明直接型AIC实现结构更适合跳频信号的高效采集；第四部分提出了三种跳频压缩采样中的定频抑制技术，并构建了基于精简测量矩阵的跳频压缩采样结构模型，通过仿真验证了该结构不仅可以实现定频抑制，又能高概率的完成跳频信号的重建。更多还原

【Abstract】 Since the frequency hopping technology was invented, it has been widely used in militarycommunications because of the advantages of anti-jamming, low probability, high spectralefficiency and so on. Howerver, the frequency hopping bandwidth becomes wider and wider.Because of the limitation of the Nyquist sampling theory, we have encountered many difficulties onaccount of the level of existing hardware technology in the applications of ultra high frequencyhopping communication system.The compressive sampling theory appeared these years breaks through the limitation of theNyquist sampling theory. The compressive sampling theory points out that we can efficientlysample signals by using few of the incoherence measurements with low speed if the signals aresparse in some transform domain. Corresponding to the difficulties encountered in the ultrawideband frequency hopping signal detection recently, we do research into the AIC technology offrequency hopping using compressive sampling when we apply compressive sampling to thefrequency hopping signal acquisition. However, the large wideband of the frequency hoppingsignals will definitely lead to the mix of numerous of fixed-frequency interference signals, whichincrease the sparsity of the frequency hopping signal and influence the efficiency of thecompressive sampling of frequency hopping signal. In order to solve the hardware and algorithmproblems caused by the direct compressive sampling which obeys the rule“Reconstruction first,then inhibition”, we proposed three fixed-frequency inhibition methods which obey the rule“Inhibition first, then reconstruction”. The first one is posteriori compressive sampling. The secondone is annihilation filter. The third one is the method of using minimum mean square error. Andaccording to these three methods of fixed-frequency inhibition, we constructed a model based thesimplified measurement matrix of the compressive sampling of frequency hopping signal, whichgave us a new idea to solve the acquisition of ultra wideband frequency hopping signal.There are four main parts in this paper. The first part introduces the basic principles offrequency hopping technology as well as the method of the power spectrum elimination applied tofixed-frequency inhibition in traditional detection. The second part introduces compressivesampling theory including sparsity expression of signal, design of measurement matrix andreconstruction, then we verify the feasibility of the theory by simulation. The third part introducestwo typical implementation structures of AIC: direct type and pre-modulation. According to thecharacteristics of ultra wideband frequency hopping signal characteristics, the simulation resultsshow that direct type is more suitable for the frequency hopping signal efficiently acquisition. The fourth part proposed three methods of fixed-frequency inhibition in the compressive sampling offrequency hopping signal and constructed the fixed-frequency inhibition model of the compressivesampling of frequency hopping signal. Through the simulation resultes, we proved that the structurenot only can realize fixed-frequency inhibition, but also can finish the reconstruction of frequencyhopping signal with high probability.更多还原