Events

Title: Generative AI and Deep Learning for Resource Allocation in 6G Wireless NetworksAbstract:This talk provides an in-depth exploration of resource management in 6G wireless networks, focusing on the vision, key performance indicators (KPIs), key enabling techniques (KETs), and the diverse array of services characteristic of these advanced networks. The distinct challenges inherent in 6G resource management call for a pivotal shift toward artificial intelligence (AI) and machine learning (ML)–driven solutions, requiring a departure from traditional optimization-centric <a href="http://approaches.The" target="_blank" title="approaches.The">approaches.The talk sheds light on generative AI and unsupervised ML strategies tailored to effectively address convex and non-convex resource management optimization problems. A key focus is placed on deep unsupervised learning techniques for network resource allocation under nonlinear and non-convex constraints. Deep implicit layers and differentiable projection methods are explored as mechanisms to ensure zero constraint violations in applications such as beamforming, phase-shift optimization, and power <a href="http://allocation.Furthermore" target="_blank" title="allocation.Furthermore">allocation.Furthermore, the potential of generative AI models, including large language models (LLMs), to enable proactive network resource allocation is examined, highlighting their role in optimizing performance and reducing reliance on traditional heuristics. The session concludes by identifying key research gaps and future directions, paving the way for next-generation AI-driven wireless <a href="http://networks.Co-sponsored" target="_blank" title="networks.Co-sponsored">networks.Co-sponsored by: Hong-chuan YangRoom: 660, Bldg: Engineering/Computer Science Building (ECS), 3800 Finnerty Road, Victoria, British Columbia, Canada, V8P 5C2