A Cognitive Routing framework for Self-Organised Knowledge Defined Networks

This study investigates the applicability of machine learning methods to the routing protocols for achieving rapid convergence in self-organized knowledge-defined networks. The research explores the constituents of the Self-Organized Networking (SON) paradigm for 5G and beyond, aiming to design a routing protocol that complies with the SON requirements. Further, it also exploits a contemporary discipline called Knowledge-Defined Networking (KDN) to extend the routing capability by calculating the “Most Reliable” path than the shortest one.
The research identifies the potential key areas and possible techniques to meet the objectives by surveying the state-of-the-art of the relevant fields, such as QoS aware routing, Hybrid SDN architectures, intelligent routing models, and service migration techniques. The design phase focuses primarily on the mathematical modelling of the routing problem and approaches the solution by optimizing at the structural level. The work contributes Stochastic Temporal Edge Normalization (STEN) technique which fuses link and node utilization for cost calculation; MRoute, a hybrid routing algorithm for SDN that leverages STEN to provide constant-time convergence; Most Reliable Route First (MRRF) that uses a Recurrent Neural Network (RNN) to approximate route-reliability as the metric of MRRF. Additionally, the research outcomes include a cross-platform SDN Integration framework (SDN-SIM) and a secure migration technique for containerized services in a Multi-access Edge Computing
environment using Distributed Ledger Technology.
The research work now eyes the development of 6G standards and its compliance with Industry-5.0 for enhancing the abilities of the present outcomes in the light of Deep Reinforcement Learning and Quantum Computing.