The recent developments in the field of edge computing and artificial intelligence (AI) highlight a significant shift towards optimizing AI models for deployment on resource-constrained edge devices. This trend is driven by the need to process data locally, reducing latency and bandwidth usage, and enhancing privacy and security. Innovations in data, model, and system optimization strategies are at the forefront, enabling more efficient and reliable edge AI applications. Additionally, the integration of microservice architectures into edge networks is being explored to improve service flexibility and scalability, despite the challenges posed by complex network topologies and dispersed node locations. Energy efficiency remains a critical concern, with new approaches to resource allocation and computation offloading being developed to sustain continuous communication in energy-harvesting powered wireless sensor networks. Furthermore, the emergence of semantic communication and collaborative beamforming techniques is addressing the challenges of data redundancy and energy management in IoT networks. The application of deep learning to IoT and the development of dynamic scheduling and migration frameworks for cloud-assisted edge clusters are also notable advancements, aiming to optimize resource utilization and ensure quality of service for latency-sensitive applications. Lastly, the integration of contract-inspired contest theory with deep reinforcement learning and generative diffusion models is paving the way for high-quality, photorealistic image generation in mobile edge metaverse applications, enhancing user experience in immersive virtual environments.

Noteworthy Papers

• Optimizing Edge AI: A Comprehensive Survey on Data, Model, and System Strategies: Proposes an optimization triad for efficient edge AI deployment, addressing data, model, and system-level challenges.
• Resource Allocation for the Training of Image Semantic Communication Networks: Introduces a distributed system and a resource allocation algorithm to optimize training time and energy consumption for image semantic communication models.
• Topology-aware Microservice Architecture in Edge Networks: Deployment Optimization and Implementation: Presents a topology-aware microservices deployment scheme that significantly improves communication delay and network performance.
• Towards Applying Deep Learning to The Internet of Things: A Model and A Framework: Offers a DL optimization model and a framework to ease the selection and re-use of DLNs on IoTs, enhancing performance without sacrificing quality.
• Energy Efficient Computation Offloading and Virtual Connection Control in Uplink Small Cell Networks: Develops a joint resource allocation, energy efficiency, and flow control algorithm to solve nonconvex and hierarchical optimization problems in HetNet systems.
• A Correlated Data-Driven Collaborative Beamforming Approach for Energy-efficient IoT Data Transmission: Introduces a novel communication framework integrating collaborative beamforming with an overlap-based multi-hop routing protocol to enhance data transmission efficiency.
• Energy-Aware Resource Allocation for Energy Harvesting Powered Wireless Sensor Nodes: Proposes an energy-aware resource allocation problem and an iterative algorithm to maximize long-term system throughput in EH-powered wireless sensor networks.
• KubeDSM: A Kubernetes-based Dynamic Scheduling and Migration Framework for Cloud-Assisted Edge Clusters: Introduces KubeDSM, a framework for dynamic scheduling and migration in cloud-assisted edge environments, improving resource utilization and QoS for latency-sensitive applications.
• Contract-Inspired Contest Theory for Controllable Image Generation in Mobile Edge Metaverse: Presents a framework combining contract-inspired contest theory, DRL, and GDMs to optimize image generation in resource-constrained mobile edge computing environments.