International Conference on Future and Intelligent Networking (FINE 2026)

Aug 26-28, 2026 · Osaka, Japan.

http://www.ieee-fine.org/2026/

Track 6: Internet of AI Agents, Embodied AI Networks, and Network for AI

Track Chairs:
Stefano Salsano, University of Rome Tor Vergata, Italy stefano.salsano@uniroma2.it
Zhiqing Tang, Beijing Normal University, China zhiqingtang@bnu.edu.cn
Nirmala Shenoy, Rochester Institute of Technology, USA nxsvks@rit.edu

Description

Artificial intelligence is rapidly evolving from centralized cloud-based systems to distributed, agentic, and embodied ecosystems operating across data centers, edge environments, and physical worlds. The emerging Internet of AI Agents and Embodied AI Networks require rethinking foundational networking architectures, protocols, and control planes to support scalable, low-latency, secure, and energy-efficient AI workloads.

This track invites original research contributions addressing networking foundations for agentic and embodied AI systems, as well as network architectures and protocols optimized for AI workloads in data centers, edge environments, and cyber-physical systems.

We welcome submissions spanning theoretical foundations, system design, experimental evaluations, and real-world deployments.

The convergence of AI, networking, and distributed systems presents an inflection point comparable to the early Internet era. Supporting agentic and embodied AI at scale demands new abstractions beyond traditional IP-centric models, enabling autonomous, secure, and high-performance AI-native networks.

We invite the community to contribute foundational ideas, practical systems, and bold new architectures shaping the future Internet of AI systems.

Track Topics

• Internet of AI Agents
• Research on networking architectures and protocols enabling distributed AI agents to communicate, coordinate, and operate across domains.
• Agent-centric network architectures
• Naming, addressing, and identity frameworks for AI agents
• Secure inter-agent communication protocols
• Multi-agent coordination and distributed consensus
• Cross-domain agent routing and federation
• Trust, governance, and policy enforcement for autonomous agents
• Semantic and intent-aware networking for AI-to-AI interaction
• Observability and explainability in agent communication
• Edge–cloud orchestration of distributed agents
• Scalability and performance optimization for large agent ecosystems
• Embodied AI Networks and Digital–Physical Interaction
• Research addressing communication, coordination, and reliability challenges in robotic, swarm, and cyber-physical AI systems.
• Swarm communication protocols
• Robot and drone mesh networking
• MANETs for embodied AI systems
• Ultra-reliable low-latency communication (URLLC)
• Deterministic and time-sensitive networking (TSN)
• Digital twin synchronization and cyber-physical feedback loops
• Distributed sensor fusion over networks
• Resilient AI coordination under network partition
• Energy-efficient communication for mobile and embedded AI
• Security and safety mechanisms for embodied AI networks
• Networks for AI Workloads
• Research on network architectures and protocols supporting large-scale AI training and inference.
• AI-optimized data center network architectures (e.g., Clos/leaf–spine fabrics)
• Congestion control for AI workloads
• Distributed training and inference networking
• AllReduce and gradient aggregation optimization
• RDMA, RoCE, and high-performance interconnects
• In-network computing and programmable data planes
• SmartNICs and network offload for AI
• Optical and memory-centric AI fabrics
• CDN and network cache allocation for AI model distribution
• Model-aware routing and inference placement
• Scalable interconnects for GPU clusters
• Fault tolerance and fast convergence in AI clusters
• Cross-Cutting Themes
• We particularly encourage submissions exploring:
• Clean-slate networking architectures for AI-native systems
• Energy-efficient and carbon-aware AI networking
• AI-assisted autonomous network control
• Security and trust in AI control planes
• Network-aware AI workload placement
• Formal analysis of stability and convergence in AI fabrics
• Scalable inter-domain architectures for AI ecosystems

Paper Submission and Publication

Details of paper submission and publication can be found here.