1. 1 USC INFORMATION SCIENCES INSTITUTE RAP: An End-to-End Congestion Control Mechanism for Realtime Streams in the Internet Reza Rejaie, Mark Handley, Deborah Estrin Scalable Personal Teleconferencing http://netweb.usc.edu/reza/rap.html INFOCOM’99 March 25, 1999

2. 2 USC INFORMATION SCIENCES INSTITUTE Motivation n Rapid growth in deployment of streaming applications(audio, video) over the Internet n TCP is inappropriate for realtime streams n A well-behaved congestion-controlled UDP is desired

3. 3 USC INFORMATION SCIENCES INSTITUTE Best-effort Networks(The Internet) n Shared environment Š Available bandwidth is not known a priori Š Bandwidth can change widely in a session Š Seemingly-random losses n TCP-based traffic dominates n End-to-end congestion control is crucial for stability, fairness & high utilization n The main requirement in the Internet is to be “well-behaved” & in particular “TCP-friendly”

4. 4 USC INFORMATION SCIENCES INSTITUTE Previous works n Modified TCP Š [Jacob et al., 97], SCP[Cen et al., 98] n TCP equation Š [Padhye et al., 98] n Additive Inc., Multiplicative Dec. Š LDA[Sisam et al., 98] n Challenge: TCP is a moving target

5. 5 USC INFORMATION SCIENCES INSTITUTE Overview of RAP n Decision Function n Increase/Decrease Algorithm n Decision Frequency n Goal: to be TCP-friendly Time Rate Decision Frequency Decision Function + -- Increase/Decrease Algorithm

6. 6 USC INFORMATION SCIENCES INSTITUTE Congestion Control Mechanism n Adjust the rate once per round-trip- time (RTT) n Increase the rate periodically if no congestion n Decrease the rate when congestion occurs Š Packet loss signals congestion n Cluster Loss Š Grouping losses per RTT

7. 7 USC INFORMATION SCIENCES INSTITUTE Rate Adaptation Algorithm n Coarse-grain rate adaptation Š Additive Increase, Multiplicative Decrease (AIMD) n Fine-grain rate adaptation Š Increase responsiveness to transient congestion Š Emulates congestion avoidance mechanism in TCP Š The ratio of short-term to long-term average RTT

8. 8 USC INFORMATION SCIENCES INSTITUTE Coarse vs fine grain RAP fig Impact of fine-grain rate adaptation

9. 9 USC INFORMATION SCIENCES INSTITUTE Evaluation Challenges n Large number of parameters, e.g. BW, RTT, No of flows, background traffic, TCP type,... n Dependency among parameters n TCP is a moving target! Ô An evaluation methodology is required

10. 10 USC INFORMATION SCIENCES INSTITUTE n RAP against Tahoe, Reno, NewReno & SACK n Measuring the goodput n Fairness ratio n Configuration parameters : Š Bandwidth per flow Š RTT Š Number of flows RAP Sinks TCP Sinks RAP Traffic SW TCP Traffic SW RAP Sources TCP Sources Simulation

11. 11 USC INFORMATION SCIENCES INSTITUTE Fairness ratio across the parameter space without F.G. adaptation

12. 12 USC INFORMATION SCIENCES INSTITUTE Fairness ratio across the parameter space with F.G. adaptation

13. 13 USC INFORMATION SCIENCES INSTITUTE Impact of RED switches on Fairness ratio

14. 14 USC INFORMATION SCIENCES INSTITUTE Conclusion n RAP achieves TCP-friendliness over a wide range Š Fine grain rate adaptation extends inter-protocol fairness to a wider range n Occasional unfairness against TCP traffic is mainly due to divergence of TCP congestion control from AIMD Š Pronounced more clearly for Reno and Tahoe Š The bigger TCP’s congestion window, the closer its behavior to AIMD n RED gateways can improve inter-protocol sharing Š Depending on how well RED is configured

15. 15 USC INFORMATION SCIENCES INSTITUTE Future Work n RAP’s behavior in the presence of short-lived TCP traffic n Deploying RAP for unicast delivery of realtime streams n Using RAP for multicast congestion control in semi-homogeneous group < http://netweb.usc.edu/reza/rap.html