Delay-based Congestion Control for Multipath TCP Yu Cao, Mingwei Xu, Xiaoming Fu Tsinghua University University of Goettingen Outline Background and problem statement Congestion Equality Principle Weighted Vegas Simulations Conclusions Multipath Transfer Ever-increasing multihomed hosts Split traffic across multiple paths. Provide a new opportunity for designers to enhance performance of end-to-end transmission. Benefits for End-hosts Increase throughput Improve robustness [MPTCP, NSDI 2011] WiFi: high rate unstable low coverage 3G: low rate stable high coverage Benefits for networks Bandwidth can be more fairly and efficiently shared by flows. S1 S2 S3 6M 9M D1 D2 D3 New Requirements for MPCC Improvement on throughput is constrained by fairness. S1 S2 S3 9M D1 D2 9M 9M D3 Traffic engineering at end-systems Coupling Subflows Together Regard network resources as a whole to compete for bandwidth S1 S2 S3 6M 9M D1 D2 D3 S1 S2 D1 15M D2 S3 D3 How to determine appropriate rates on each path? How to shift traffic with only local knowledge for sources ? Congestion Equality Principle A fair and efficient traffic shifting implies that every flow strives to equalize the degree of congestion that it perceives on all its available paths. A knob to control rates A metric to estimate congestion degree Delay-based vs. Loss-based Packet queuing delay Packet loss events Multi-bit info quantifing congestion degree Single-bit congestion signals Be sensitive to … Perceive changes of congestion in a large timescale RTT fairness Bias against large-RTT flows Low buffer consumption Frequent losses Less aggressively More aggressively -- Linked Increases, CMT/RP Understanding TCP-Vegas cwnd rtt baseRTT cwnd cwnd diff baseRTT rtt baseRTT , cwnd 1 diff , cwnd 1 cwnd x rtt diff q rtt baseRTT The number of backlogged packets x q Bandwidth Sharing 3 2 1 3M 6M 2M 1M Weighted Vegas Core algo.: allocate alpha to each subflow. 6M 5M 15M 5 5 5 5M 5M 1:4 9M 1M 4M 5M 5M 1:4 ? 5 1 5 4 6 9 To equalize congestion degree of the two paths. 5 5 6 9 Network Utility Maximization max U s ( ys ) min Ls (λ ) λcT s.t. y Bx Ls (λ ) : max U s xs , r qr xs , r xs ,r 0, rRs rRs rRs x0 s S Ax c λ 0 sS s Given a fixed budget, invest it in the cheapest paths to maximize the utility. Lowest queuing delay Iteratively Tweaking Weights The total amount of backlogged packets is fixed at s , regardless of the number of subflows. s , r (t ) ks , r (t 1) s Control rates xs , r (t ) s , r (t ) qr xs , r (t ) xs , r (t 1) 1 Update parameters ks , r (t ) xs , r (t ) x iRs s ,i (t ) Tweak weights A summary of weighted Vegas Runs in the same way as TCP-Vegas on each path. s Uses equilibrium rates of subflows to adjust weights. is allocated to subflows according to weights. A larger s means more packets are backlogged in link queues. A quite small s makes wVegas over sensitive to the noise of RTT. Simulations We implemented wVegas and Linked Increases in NS-3. Focus on the fairness and efficiency Expect wVegas achieves a fine-grained traffic shifting. s 10 Iteratively adjust rate Two bottleneck links Transmission rate wVegas Linked Increases Fairness on Bottleneck Links wVegas Linked Increases Dynamics of Traffic Shifting The Domino Effect Rate complementation between subflows Conclusions The Congestion Equality Principle wVegas can achieve fine-grained traffic shifting. wVegas relies on the accurate measurement of RTTs. wVegas and Linked Increases have their own respective advantages and defects. Combine they two together? Thanks
© Copyright 2026 Paperzz