학술논문

Customizing service path based on polymorphic routing model in future networks
Document Type
Periodical
Source
China Communications China Commun. Communications, China. 16(7):207-228 Jul, 2019
Subject
Communication, Networking and Broadcast Technologies
Routing
Internet
Adaptation models
Security
Atomic layer deposition
Network architecture
Adaptive systems
polymorphic
base state
routing model
MDP
reinforcement learning
Language
ISSN
1673-5447
Abstract
The current Internet has evolved during the last decade to a global provider of diverse applications. However, the underlying structure of routing and addressing has not evolved in the same pace and is somewhat inflexible. How to provide diverse routing services, support emerging communication paradigms based on limited and definite network resources has become an urgent challenge. This paper investigates the adaptive matching between routing and application through network function decomposition and composition, and proposes a polymorphic routing model to support diverse applications and emerging communication paradigms. The model splits complex routing functions into its constituents, and derives customized routing mechanisms supporting various applications by composing the routing constituents. The derivation process is modeled as a Markov Decision Process (MDP), and a polymorphic derivation algorithm is also proposed to derive customized routing instances for diverse applications. The model enables the network to self-adjust routing services dynamically to adapt to the different requirements of applications, supports coexistence of multiple routing modes and communication paradigms, and provides a feasible solution for the network compatibility and evolvement. We describe the key design and demonstrate the feasibility of polymorphic derivation by simulations. We also present case studies that demonstrate key functionalities the polymorphic routing model enables.