EAGER: Spectrum Situational Awareness-Understanding the Data

点击翻译按钮获取中文摘要

The key resource shared by all wireless systems is the electromagnetic spectrum, which is challenged by temporal and spatial congestion, interference, and increasing security threats. This project investigates the hypothesis that building a resilient always-on wireless network, capable of meeting ever-increasing performance and security requirements, necessitates not only innovations in radio design but also innovations in the management and exploitation of vast amounts of radio spectrum knowledge. Such knowledge is referred to as spectrum situational awareness - the amount of actionable intelligence and understanding that a wireless network (or individual radio) has about its RF spectrum environment. This research work formally studies the nature of big spectrum data and investigates several critical issues, e.g., how should data from heterogeneous sources be prioritized or weighted, and what are the effects of inaccurate or sparse data measurements.The scientific goals of this project are to provide foundational principles and insights for developing innovative and scalable spectrum situational awareness architectures, and to enable more efficient and effective uses of the data as well as the improved protocol and system designs. Using formal statistical methods (e.g., regression analysis and response surface methodology) and new big data analytics tools, the research team fully characterizes big spectrum data, its potential impact on spectrum utilization and system performance, and its interaction with the plethora of influential radio, network, and environmental factors. The empirical work is based on experimental scenarios that reflect realistic networking environments, comprising multiple wireless access technologies, interconnected high-speed backhaul networks, and distributed cloud-based technologies. Two testbed instruments are used to design and execute large-scale simulation and emulation experiments: the Distributed Computing and Visual Analytics Sandbox for High Volume Data Streams combined with the Wireless Systems and Performance Engineering Research (WiSPER) laboratory. Additionally, the research team develops a multilayer spectrum situational awareness framework and is deriving performance indices that can be used to quickly quantify and compare the spectrum situational awareness capabilities of network systems. The project engages students from underrepresented groups via ongoing partnerships with Historically Black Colleges and Universities.