|
Abstract:
|
In the context of modern high-speed communication networks, decision reactivity is often complicated by the notion of guaranteed Quality of Service (QoS), which can either be related to time, packet loss or bandwidth requirements: constraints related to various types of QoS make some algorithms not acceptable. Due to emerging real-time and multimedia applications, efficient routing of information packets in dynamically changing communication network requires that as the load levels, traffic patterns and topology of the network change, the decision policy also adapts. We focused in this paper on QoS based mechanisms by developing a neuro-dynamic programming to construct dynamic state-dependent policies. In this paper, we present an accurate description of the current state- of-the-art and give an overview of our work in the use of reinforcement learning concepts focused on communications networks. We focus our attention by developing a system based on this paradigm and study the use of reinforcement learning approaches in three different communication networking domains: wired networks, mobile ad hoc networks, and packet router’s scheduling networks. |