Compressed Localization and Spectrum Sensing for Cognitive Radio and Distributed Radio Surveillance (CLASS)

宽带无线电发射的检测、识别和定位是射频工程和通信领域中一个非常相关且具有挑战性的研究课题。根据国际电信联盟（ITU）的规则，无线电监测是代表政府当局进行频谱管制和管理，以监测所分配和管制良好的频带的法律的使用。重要性甚至将随着未来的频率捷变认知无线电接入系统而增加，该系统将对合法分配但空闲的频带进行二次利用。最近，无线电监视方法也越来越多的相关性执法，公共安全，救援工作等，我们认为空间分布的无线电观测节点，相互连接，并通过通信链路连接到数据融合中心的数据收集和节点的命令和控制的监视网络。该节点具有快速宽带频谱-时间感知、辐射源检测和识别以及与位置相关的参数估计（如到达方向和时间差以及接收信号强度）的功能。通过数据融合和地理定位方法，获得了某一明确区域内无线电辐射的天气视图。主要的研究影响来自压缩感知方法（CS）的应用。CS是一个数学框架，允许有效的采样稀疏问题和重建欠定方程。这适用于这里，因为无线电发射在占用频率、接入时间、地理位置以及调制格式方面是稀疏的。与传统的Nyquist感知算法相比，CS算法能够显著减少相关数据量，更适合于观测节点与融合中心之间有限数据速率的通信和协作。这避免了信息丢失，这将损害传感器数据融合。在一般情况下，CS需要在传感功能设计的范式转变，这将导致宽带天线阵列接收机，位置估计，并为合作数据采集全新的架构。进一步的创新源于严格包括多径传播效应，其不应仅被解释为必须减轻的负担。相反，我们将利用多径作为一种手段，应用先验知识的环境，因此，以提高最终的估计。该项目嵌入在德国-哥伦比亚电气工程合作研究倡议（GeCoCo-EE）的框架内，该框架基于德国Deutsche Forschungsgemeinschaft e. V.（DFG）和哥伦比亚Departamento Administrativo de Ciencia，Tecnologia e Innovación（COLCIENCIAS）之间的谅解备忘录（MoU）。

Detection, identification, and localization of wideband radio emissions is a very relevant and challenging research topic in RF engineering and communications. According to the rules of the International Telecommunication Union (ITU), radio surveillance is carried out on behalf of governmental authorities for spectrum regulation and management to monitor legal usage of the allocated and well regulated frequency bands. Importance will even increase with future frequency agile cognitive radio access systems that will make secondary usage of legally allocated, but idle bands. Recently, radio surveillance methods are gaining increased relevance also for law enforcement, public safety, rescue work, etc. We consider a surveillance network of spatially distributed radio observer nodes that are mutually connected and linked to a data fusion center by communication links for data collection and node command and control. The nodes feature fast wideband spectral-temporal sensing, emitter detection and identification, and location related parameter estimation such as direction and time difference of arrival as well as received signal strength. A synoptic view of radio emissions within a certain well defined area is gained by data fusion and geolocalization methods. The major research impact comes from the application of compressed sensing methods (CS). CS is a mathematical framework that allows efficient sampling of sparse problems and reconstruction from underdetermined equations. This applies here since the radio emissions are sparse in occupied frequency, access time, geolocation, and also in terms of the modulation format. Compared to conventional Nyquist sensing, CS methods support a considerable reduction of the relevant data volume, which suits better for limited data rate communication and cooperation between the observer nodes and the fusion center. This avoids loss of information, which would compromise sensor data fusion. In general, CS requires a paradigm shift in sensing function design which will lead to completely new architectures for wideband antenna array receivers, location estimation, and for cooperative data acquisition. Further innovation results from the strict inclusion of multipath propagation effects that shall not only be interpreted as a burden which has to be mitigated. Instead, we will exploit multipath as a means of applying a priori knowledge about the environment and, hence, to enhance the final estimation. This project is embedded within the Framework of the German-Colombian Collaborative Research Initiative in Electrical Engineering (GeCoCo-EE), which is based on the Memorandum of Understanding (MoU) between Deutsche Forschungsgemeinschaft e.V. (DFG), Germany and Departamento Administrativo de Ciencia, Tecnologia e Innovación (COLCIENCIAS), Colombia.