A policy-based framework for autonomic reconfiguration management in heterogeneous networks

This paper presents a policy-based framework to approach the issue of autonomous reconfiguration management in heterogeneous networks. In contrast to existing policy-based approaches, the proposed framework addresses the management issue from a new perspective through posing it as a problem of learning from current network behavior, while creating and updating policies dynamically in response to changing reconfiguration requirements, and this task is implemented by Reinforcement Learning methodology. A two-layer policy model is used to mapping users and operators' higher level goals into network level objectives. The autonomic reconfiguration procedures for policy creation, storage, evaluation are also presented in detail. Illustrative examples analysis and simulation results demonstrate the performance of the proposed work.