Dynamic Application Profiles using Neural Networks for adaptive quality of service support in the Internet

The Internet continues to evolve into a network that supports a wide spectrum of applications with diverse requirements. The introduction of multimedia poses delay, throughput and packet loss requirements towards the network and makes the provision of QoS a very important issue. In this paper, we present the concept of Dynamic Application Profiles for the transparent support of QoS in the Internet. The paper builds on existing work on a distributed QoS architecture, and uses intelligent mechanisms, such as Neural Networks, in an effort to provide dynamic QoS to network applications in a fashion comprehensible to the average end user.