Dynamic-CBT and ChIPS – Router Support for Improved Multimedia Performance on the Internet Jae Chung and Mark Claypool Computer Science Department Worcester Polytechnic Institute Worcester, Massachusetts, USA

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

Multimedia Using TCP

Multimedia Using UDP

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

Router Queue Management Drop Tail (FIFO) Resource Reservation Active Queue Mgmt... (Floyd, Jacobson ‘95) CBQ... (Floyd, Jacobson ‘93) RED Class-Based Threshold (CBT) Class-Based Threshold (CBT) (Parris, Jeffay, Smith ‘99) Bursty Loss Unfairness

Outline  Introduction  CBT   D-CBT  ChIPS  Conclusion

CBT Design

Router Queue Management Drop Tail (FIFO) Resource Reservation... Active Queue Mgmt... (Lin, Morris ‘97) FRED Class-Based Threshold (CBT) Class-Based Threshold (CBT) (Floyd, Jacobson ‘95) CBQ (Floyd, Jacobson ‘93) RED (Parris, Jeffay, Smith ‘99) ChIPS Dynamic-CBT

D-CBT Design Responsive MM Enable Thrsh

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

Simulation (RED, CBT, D-CBT) 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 = max_prb = 0.1 CBT: mmu_th = 2.9 udp_th = 0.6 Second

Fairness: RED

Fairness: CBT

Fairness: D-CBT

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

Cut-In Packet Scheduling (ChIPS) Design

ChIPS Evaluation - Jitter

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

Conclusion  D-CBT –Protect responsive flows from unresponsive –Class-based consideration –Adaptive to mix of flows  ChIPS –Reduce multimedia jitter during congestion

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

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