Providing End-to-End QoS for IP-based Latency-sensitive Applications Dissertation Proposal

Advances in IP-technology, which include the real-time encoding and decoding of voice and vid router mechanisms, and real-time transport protocol (RTP) offer great potential for computer-teleph integration over IP networks. However, due to the unpredictable loss, delay, and delay jitter in pack switching network, the conventional Internet single best-effort class of service can adversely impac end-to-end performance of real-time applications. This dissertation proposal describes a research e determining mechanisms to provide Quality of Service (QoS) in an efficient and scalable manner ov Internet. We will elaborate on an end-to-end service architecture that can support per session statis quality of service guarantees, by building on the significant amount of previous research on networ application-level QoS solutions. In our initial work, we have chosen Voice over IP (VoIP) traffic as a w load for evaluation, and use a subjective test to determine the effects of network centric parameters human perceived voice quality. First, we develop resource allocation techniques for VoIP traffic in virtual private networks (VPN achieve the required QoS. Specifically, we will show how to use traffic statistics in predicting aggreg bandwidth usage, which is crucial for capacity planning and admission control decisions. Since the results may not scale to the publicly shared Internet, we propose to study performance of real-time a tions within the Diff-Serv framework. We will describe a) the trade-offs between performance and net efficiency for VoIP traffic in the presence of other integrated computer-telephony applications, e.g., active web applications, and b) the interaction between real time flows transported using UDP, and c tion-controlled TCP connections. We expect to make modifications to the existing RTP protocol, so real-time applications will adapt their rates to network load, and prevent potential starvation of TCP nections. Finally, as an example, we will investigate the role of adaptive audio applications by re-ev ing the play-out delay adjustment algorithms in the modern context. We have shown the existence correlation between packet delays, and propose to design and analyze a delay adjustment algorith on this conditional delay estimate. We will evaluate this algorithm using trace-based simulations.