-
Design objective:
•Assemble a protocol to take a wide range of control functions which can self-evolve and transform for:
•Different applications with heterogeneous service requirements
•*-cast service modes
•Ubiquitous coverage
•Ubiquitous intelligence
Data plane and control plane cubic model:
•User control: user-associated billing, accounting, …
•Session control: session-level authentication, policy rules, mobility management, …
•Flow/packet control: Flow/packet-level QoS support
User and session control can be in mobile core (cloud), while flow/packet control should follow data plane.
New protocol architecture for Self-Evolving and Transformative (SET) Intelligent Protocols
- Modularity: decompose protocol control functions so a protocol can be assembled and configured using different control algorithms and configurations.
2. Protocol Control Agent (PCA): In each networking entity, its PCA interacts with its full protocol stack. E.g., in the end system, PCA can receive the user’s QoS requirements, transport layer service quality measurements, and network/link measurements, as well as user location, mobility, etc. Then it relies on various online/offline machine learning and AI algorithm to make intelligent decision in assembling and configuring the protocols in each entity.
3. Protocol configuration and adaptation: PCA uses default configurations for easy-start, and update them based on the real-time measurements and feedback.
[1] L. Cai, J. Pan, W. Yang, X. Ren, and X. Shen, “Self-Evolving and Transformative (SET) Protocol Architecture for 6G,” IEEE Wireless Communications Magazine, 30(4): 178-186, August 2023.