The field of edge computing and AI-driven resource optimization is rapidly advancing, with a focus on developing innovative solutions to improve the efficiency, scalability, and sustainability of computing systems. Researchers are exploring the use of deep reinforcement learning, multi-agent systems, and hierarchical co-optimization frameworks to optimize resource allocation, task scheduling, and power dispatch in edge computing environments. These approaches aim to reduce latency, energy consumption, and costs while ensuring high-quality services and strict Service Level Agreement (SLA) compliance. Noteworthy papers in this area include:

• A novel deep reinforcement learning-based QoS-aware LLM routing framework, which significantly improves average QoS and computing resource efficiency.
• A Wireless Multi-Agent System (WMAS) that provides intelligent and customized services for different user equipment, achieving higher task performance and lower conversation overhead.
• An adaptive AI agent placement and migration framework, which reduces deployment latency and migration costs in dynamic edge environments.
• A SLA-aware multi-objective reinforcement learning framework (SLA-MORL) that intelligently allocates GPU and CPU resources, achieving significant reductions in training time, costs, and SLA violations.
• A novel hierarchical co-optimization framework for coordinated task scheduling and power dispatch in Computing Power Networks, which reduces total carbon emissions and operational costs while maintaining stringent QoS for computational tasks.