基于多维参数感知与行为学习的无线分布式网络多源MIMO混叠信号识别

As an important key technology, signal-parameter identification has been widely applied in many realms, such as cognitive radio, eavesdropping, countermeasure and civilian security-monitoring. With the deployment of multi-antenna terminals in wireless distributed networks (WDN), the eavesdropper will receive MIMO overlapped signals from multiple nodes. Consequently, antenna-grouping ambiguity arises and leads to a new challenge in signal parameter identification for WDNs. .In this work, by investigating the overlapped mechanism of multi-source MIMO signals, we implement an information fusion of the network behavior and node-signal characteristic. Further, a novel multi-dimension parameter sensing and behavior learning based approach to identify the multi-source overlapped MIMO signals is proposed. The following three major problems are well investigated in this work. 1) State-variation sensing for multi-user asynchronous signals in WDNs. 2) Multi-dimension characteristics integrated-extraction of virtual MIMO signals with low detectable probability. 3) Screening and classification of combinational behavior constructed from the network behavior and signal characteristics..By eliminating the overlap among multi-source MIMO signals, the identification performance of the eavesdropper will be improved to a large extent. The outcome is applicable to many engineering realms, as well as establishes a theoretical basis for the intelligent receiver design in cognitive radios, direction-finding and positioning of radiation source, and capture of the information dominance.

无线分布式网络的信号参数识别是认知无线电、侦听对抗、民用安防等领域的关键技术。随着越来越多的多天线终端设备出现在无线分布式网络中，识别设备会收到自多个节点的MIMO混叠信号，导致“天线分组模糊度”的出现，给无线分布式网络信号参数识别带来新的挑战。.课题在分析多源MIMO信号混叠机理的基础上，对网络行为与节点信号特征进行有效的信息融合，提出了基于多维参数感知与行为学习的多源MIMO混叠信号识别新方法。具体研究包括：1）无线分布式网络多用户异步信号状态变化感知方法；2）低可检测环境下虚拟MIMO信号多维特征联合提取方法；3）网络行为与节点信号特征的组合行为筛选与分类。.通过有效解决多源MIMO信号混叠问题，可大幅提升网络侦听节点的识别能力。课题成果为认知无线电中的智能接收机设计、侦听对抗中提高辐射源测向定位能力、获取制信息权等应用提供重要的理论支撑。

无线分布式网络的信号参数识别是认知无线电、侦听对抗、民用安防等领域的关键技术。随着越来越多的多天线终端设备出现在无线分布式网络中，识别设备会收到自多个节点的MIMO混叠信号，导致“天线分组模糊度”的出现，给无线分布式网络信号参数识别带来新的挑战。课题在分析多源MIMO信号混叠机理的基础上，对网络行为与节点信号特征进行有效的信息融合，提出了基于多维参数感知与行为学习的多源MIMO混叠信号识别新方法。具体研究包括：.1）无线分布式网络多用户异步信号状态变化感知方法；2）低可检测环境下虚拟MIMO信号多维特征联合提取方法；3）网络行为与节点信号特征的组合行为筛选与分类。. 通过有效解决多源MIMO信号混叠问题，可大幅提升网络侦听节点的识别能力。课题成果为认知无线电中的智能接收机设计、侦听对抗中提高辐射源测向定位能力、获取制信息权等应用提供重要的理论支撑。