Modeling, Monitoring, and Optimization of Cognitive Networks

点击翻译按钮获取中文摘要

The objective of this proposal is to establish transformative novel solutions to fundamental problems in cognitive network modeling, monitoring and optimization with significant impact to both theory, algorithms, and implementation. The intellectual merit comprises the novel concept of network cartography as the tool for distilling and summarizing the "network state" for the purpose of network management. The resultant holistic approach to network monitoring offers modular and scalable statistical representations driven by past (and driving future) measurements and system design specifications. Further, it enables inference of global network behavior and adaptation of protocol designs to enhance network robustness and quality-of-service. Dramatic improvements in CR network performance will be attained by augmenting the vivid description of the network global state with models that account for the intrinsic non-tangible connectivity of cognitive wireless links, and the interference challenges arising from spectrum sharing hierarchies. The broader impacts include advances of the IEEE 802.22 compliant cognitive radio technologies, and smart grid, surveillance, geo-monitoring, and institutional networks. Furthermore, tangible advances in the fundamental tools exploited will permeate benefits to bio-informatics, speech and image compression, machine learning and data mining for social networks, and the emerging smart grid communication network protocols. Through Honors Theses and Senior Design Projects, the proposed research will also impact student training with hands-on experience in state-of-the-art wireless cognitive systems and network optimization subjects that are not fully provided by coursework.