1. Dynamic-CBT and ChIPS - Router Support for Improved Multimedia Performance on the Internet Jae Chung and Mark Claypool Department of Computer Science WPI In Proceedings of ACM Multimedia November 2000

2. The Internet and Multimedia Internet routers are best effort –No timing constraints –Packet loss, which indicates congestion TCP –Completely reliable delivery through retransmission –Respond to loss as congestion But … TCP often unsuitable for interactive multimedia –Full reliability not needed –Window-based rate fluctuations

3. Multimedia Using TCP

4. Multimedia Using UDP

5. Multimedia on the Internet Multimedia often uses UDP –Avoid delay and jitter from retransmission –Rate-based –Unresponsive! Router queue management goals –Congestion Control –Fairness –Reduce Jitter

6. Current Router Queue Management Drop Tail (FIFO) Resource Reservation... CBQ Floyd, Jacobson 93 Floyd, Jacobson 95 Lin, Morris 97 Bursty loss Unfairness (tcp, other) No QoS support RED FRED Active Queue Mgmt...

7. Class-Based Threshold (CBT) AQM Support for Multimedia - Jeffay, 99 Drop Tail (FIFO) Resource Reservation... CBQ RED FRED Class-Based Threshold (CBT) Class-Based Threshold (CBT) Active Queue Mgmt...

8. Outline Introduction CBT and D-CBT  –Design –Evaluation ChIPS –Design –Evaluation Conclusion

9. CBT Concepts CBQ + RED: Class-based isolation on RED Use Class Thresholds and Avg. # of enqueued packets on a single FIFO Queue Three classes: TCP, MM UDP (flow controlled) and Unresponsive UDP TCP MM UDP REDRED

10. CBT Design

11. CBT - Pros and Cons Pros: RED + Class-Based Isolation –Early Congestion Notification –Protect TCP, and protect (distinguish) MM UDP –Different flows coexist with predefined fairness. –Dividing bandwidth assigned is up to the class. Cons: CBQ function w/o admission control –Might not work well for certain traffic mixes –Arguable that it’s not fair (as in the case of CBQ)

12. Dynamic-CBT Drop Tail (FIFO) Resource Reservation... CBQ RED FRED Class-Based Threshold (CBT) Class-Based Threshold (CBT) Dynamic-CBT Active Queue Mgmt...

13. ChIPS Dynamic-CBT and ChIPS Drop Tail (FIFO) Resource Reservation... CBQ RED FRED Class-Based Threshold (CBT) Class-Based Threshold (CBT) Dynamic-CBT Active Queue Mgmt...

14. D-CBT Design Responsive MM Enable Thrsh

15. Flow Counting in D-CBT For every incoming packet, insert or update info and update count –Sorted Linked List - O(n) –Hash Table - O(1) Every  ms, delete old info and update count –Sorted Linked List - O(n) –Hash Table - O(n) (How are flows counted in FRED?)

16. Evaluation in NS Developed responsive multimedia application (for tagged UDP class) –AIMD Media Scaling (5 discrete rates) –“MPEG-1 like” transmission rates –[CC00a], MM-Flow Implemented and validated CBT Implemented D-CBT and measured congestion time fairness –RED vs. CBT vs. D-CBT

17. Validation of CBT on NS 02060110160 180 6 ProShare - Unresponsive MM (210Kbps each) 240 FTP-TCP 1 UDP blast (10Mbps, 1KB) RED Settings: qsize = 60 pkts max-th = 30 pkts min-th = 15 pkts qweight = 0.002 max-pro = 0.1 CBT Settings: mm-th = 10 pkts udp-th = 2 pkts (Second)

18. X axis: Seconds, Y axis: Kbyte/Sec PJS99 Experimental Results NS Simulated Results Aggregate TCP Throughput under RED (Our setup is ok, so now can check our CBT test)

19. Aggregate TCP Throughput under CBT X axis: Seconds, Y axis: Kbyte/Sec PJS99 Experimental Results NS Simulated Results

20. Simulation (RED, CBT, D-CBT) 0102030 25 FTP-TCP 30 FTP-TCP 2 CBR-UDP (5Mbps each) 10 MM-UDP s1 n1 s2 s66 s67 n2 r2 r66 r67 25Mbps, 20ms 25Mbps, 5ms n1-n2: q_size = 60 RED: max_th = 15 min_th = 5 qweight = 0.002 max_prb = 0.1 CBT: mmu_th = 2.9 udp_th = 0.6 Second

21. Jain’s Fairness Index ( f ) - Jain, 91 Examples: –1 flow –2 flows, 5 Kbps each –2 flows, 9 Kbps and 1Kbps f (x 0, x 1, x 2,…, x n ) = ( x i ) 2  i= 0 n  n xixi 2 n 0  f  1 (Greatest Fairness)

22. Fairness: RED

23. Fairness: CBT

24. Fairness: D-CBT

25. Outline Introduction CBT and D-CBT –Design –Evaluation ChIPS  –Design –Evaluation Conclusion

26. Cut-In Packet Scheduling (ChIPS) Design

27. ChIPS Evaluation - Jitter Frame Sequence Number MM Frame Delay Second D-CBT-FIFO MM1 Delay D-CBT-ChIPS MM1 Delay

28. ChIPS Evaluation - Fairness TCP Packets Delivered TCP Packet Drop Rate TCP Throughput D-CBT66,648 pkts4.46 %17,773 Kbps D-CBT w/ ChIPS 66,386 pkts4.44 %17,703 Kbps

29. Conclusion Fair to TCP Fair to Mixed (Fixed) Fair to Mixed (Variable) Improve MM Jitter D-CBT ChIPS XX X X REDX CBTXX D-CBT FIFO XX X

30. Future Work Active Flow Counting (Overhead) –For every incoming packet, update flow info +Hash Table - O(1) –Every  ms, delete old flows +Hash Table - O(n) Measure Overhead –Processing Time and Memory Usage

31. Future Work How many different classes are needed? –Example +1 class is RED +1 class per flow is FRED –Overhead per class Effects of D-CBT and ChIPS on Perceptual Quality

32. Evaluation of Science? Category of Paper Science Evaluation (1-10)? Space devoted to Experiments?