基于单通道盲分离的抗干扰技术研究

发布时间：2018-06-14 18:44

本文选题：单通道 + 抗干扰　；参考：《电子科技大学》2014年硕士论文

【摘要】：早期的扩频通信技术由于其更为理想的抗干扰能力主要应用在军事通信领域。由于民用通信需求的日益增长和频谱资源的日益稀缺,扩频通信在数字移动蜂窝通信领域也得到了广泛应用。本文针对扩频通信模型提出一种基于单通道盲分离技术的信号侦察和测量方法,并在实际环境检验了该方法的可行性。本文首先介绍了经典的盲分离算法,包括Comon算法、特征矩阵近似联合对角化(JADE)算法、不动点ICA算法(Fast-ICA),它们是单通道盲分离的理论基础。其次介绍了IS-95标准下扩频通信的信道模型,这一模型的特点决定了可以将基于ICA-R的单通道盲分离技术应用在其之上。文章的最后一部分给出了实际通信环境下的测量结果,证实了算法在抽取期望信号上的有效性以及在不同信干比(SIR)的信道环境下的抗干扰能力。由于单通道技术可以突破传统的多天线瓶颈,大大简化设备的复杂程度,因而在未来信号测量领域尤其是战术通信领域有重要实用价值。本文利用了直扩信号的特点及其在信号侦察情况下的的特殊环境。干扰信号和白噪声通常情况下和直扩信号在频域严重重叠,而传统的时频域信号处理方法很难对同频信号进行有效分离,波束成形方法也存在很多不足之处。本文提出的基于ICA-R的单通道盲信号抽取算法,可以在已知导引序列的条件下有效的分离出目的信号,显著提升信号侦察的准确程度。限于实验条件,本文利用波形贴近度和Walsh译码峰均比来考察分离效果。波形贴近度和Walsh译码峰均比表征了对待侦察的信号的估计的可靠性。仿真和实测结果表明,我们的算法得到的波形比直接采样得到的波形在贴近度上有较大改善,在信道环境较差,即低信噪比的情况下有0.1-0.2的提升。在峰均比上,也有2dB左右的性能提升。总的来说,我们可以仅利用一根接收天线即可实现提高信号侦察的可靠性。
[Abstract]:The early spread spectrum communication technology was mainly used in military communication field because of its better anti-jamming capability. Due to the increasing demand for civil communications and the increasing scarcity of spectrum resources, spread spectrum communication has been widely used in the field of digital mobile cellular communications. In this paper, a signal reconnaissance and measurement method based on single channel blind separation is proposed for spread spectrum communication model. The feasibility of this method is verified in practical environment. In this paper, we first introduce classical blind separation algorithms, including Comon algorithm, eigenmatrix approximation combined with diagonalization (JADE) algorithm, fixed point ICA algorithm (Fast-ICA), which are the theoretical basis of single-channel blind separation. Secondly, the channel model of spread spectrum communication based on IS-95 standard is introduced. The characteristics of this model determine that the ICA-R based single-channel blind separation technology can be applied to it. In the last part of the paper, the measurement results under the actual communication environment are given, and the validity of the algorithm in extracting desired signals and the anti-jamming ability in the channel environment with different signal-to-interference ratio (SIRs) are verified. Single channel technology can break through the traditional multi-antenna bottleneck and greatly simplify the complexity of the equipment, so it has important practical value in the field of signal measurement, especially in the field of tactical communication. This paper makes use of the characteristics of DS signal and its special environment in the case of signal reconnaissance. The interference signal and white noise usually overlap with the DSSS signal in the frequency domain seriously, but the traditional time-frequency domain signal processing method is very difficult to effectively separate the same frequency signal, and the beamforming method also has many shortcomings. The single channel blind signal extraction algorithm based on ICA-R proposed in this paper can effectively separate the target signal under the condition of known guidance sequence and significantly improve the accuracy of signal reconnaissance. Under experimental conditions, the separation effect is investigated by using waveform closeness and Walsh decoding peak to average ratio. The waveform closeness and Walsh decoding peak ratio represent the reliability of the estimation of the reconnaissance signal. The simulation and experimental results show that the waveform obtained by our algorithm is much better than that obtained by direct sampling in terms of closeness, and is improved by 0.1-0.2 in the case of low signal-to-noise ratio (SNR). At the peak to average ratio, there is also about 2 dB performance improvement. In general, we can improve the reliability of signal reconnaissance by using only one receiving antenna.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN914.42;TN973

【参考文献】