Dynamic Cross-Layer Design for Robust Cognitive Networking with Physical Channel and Delay Awareness

We propose to develop an operationally responsive integrated framework for cognitive radio networks with seamless user connectivity. Our key innovation is to introduce physical channel and delay awareness to dynamic spectrum access at the multi-hop networking level. In particular, we characterize the fundamental trade-offs between the throughput and delay objectives subject to realistic physical channel constraints. We leverage the cross-layer design and optimization tools at the physical, medium access control, and network layers for joint spectrum access, power control, scheduling, and routing. To that end, we integrate decentralized cognitive radio capabilities in highly congested, dynamic, multi-hop environments to meet diverse user demands. The goal is to optimize the mutual balance of link stability and capacity in spectrum-agile network operation. This involves cross-layer design with multivariate end-to-end objectives of throughput and delay over realistic physical channels. We propose a distributed framework for dynamic and predictive/reactive cognitive radio management to facilitate the adaptability across space, time, coding, and frequency dimensions optimized for network resiliency. We will evaluate the end-to-end capability of automated spectral agility and dominance of the proposed approach for cognitive cross-layer wireless networking, and compare it with static solutions of separating routing and spectrum allocation to different network layers.