Revisiting Transport Congestion Control Jian He UT Austin
![Revisiting Transport Congestion Control Jian He UT Austin 1 Revisiting Transport Congestion Control Jian He UT Austin 1](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-1.jpg)
![Why is Congestion Control necessary? Data Packets Congested Link ACK Ø Congested link vs. Why is Congestion Control necessary? Data Packets Congested Link ACK Ø Congested link vs.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-2.jpg)
![Can we distinguish congestion reasons? Ø Congestion related signals: - packet loss: duplicate ACKs, Can we distinguish congestion reasons? Ø Congestion related signals: - packet loss: duplicate ACKs,](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-3.jpg)
![Existing TCP Variants TCP Throughput-Latency Tradeoff Exploration [Remy SIGCOMM’ 13] Datacenter TCP Tail performance[TIMELY Existing TCP Variants TCP Throughput-Latency Tradeoff Exploration [Remy SIGCOMM’ 13] Datacenter TCP Tail performance[TIMELY](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-4.jpg)
![TCP Evolution Application-Specific Performance Requirements Application Sensing Layer TCP Networking Sensing Layer IP Link TCP Evolution Application-Specific Performance Requirements Application Sensing Layer TCP Networking Sensing Layer IP Link](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-5.jpg)
![Optimizing Datacenter Transport Tail Performance Mittal, Radhika, et al. "TIMELY: RTT-based congestion control for Optimizing Datacenter Transport Tail Performance Mittal, Radhika, et al. "TIMELY: RTT-based congestion control for](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-6.jpg)
![Why does tail performance matter? … Ø TCP Incast: many servers reply the client Why does tail performance matter? … Ø TCP Incast: many servers reply the client](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-7.jpg)
![Hardware Assisted RTT Measurement Why was RTT not widely used? Ø RTT-based congestion control Hardware Assisted RTT Measurement Why was RTT not widely used? Ø RTT-based congestion control](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-8.jpg)
![RTT As a Congestion Control Signal Multi-bit signal Single-bit signal Ø ECN can not RTT As a Congestion Control Signal Multi-bit signal Single-bit signal Ø ECN can not](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-9.jpg)
![RTT Correlates with Queuing Delay 10 RTT Correlates with Queuing Delay 10](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-10.jpg)
![TIMELY Framework 11 TIMELY Framework 11](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-11.jpg)
![RTT Measurement tsend Serialization Delay RTT tcompletion Propagation & Queuing Delay ACK Turnaround Time RTT Measurement tsend Serialization Delay RTT tcompletion Propagation & Queuing Delay ACK Turnaround Time](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-12.jpg)
![Transmission Rate Control Message to be sent Segments RTT Estimation Rate Controller Insert delay Transmission Rate Control Message to be sent Segments RTT Estimation Rate Controller Insert delay](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-13.jpg)
![Rate vs. Window Ø Segment size as high as 64 KB. Ø (32 us Rate vs. Window Ø Segment size as high as 64 KB. Ø (32 us](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-14.jpg)
![Rate Update 15 Rate Update 15](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-15.jpg)
![Evaluation 16 Evaluation 16](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-16.jpg)
![Datacenter Transport for Emerging Architectures Costa, Paolo, et al. "R 2 C 2: A Datacenter Transport for Emerging Architectures Costa, Paolo, et al. "R 2 C 2: A](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-17.jpg)
![Rack-Scale Computing Ø Building Block for future datacenters Ø High BW low latency network Rack-Scale Computing Ø Building Block for future datacenters Ø High BW low latency network](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-18.jpg)
![Rack-Scale Network Topology Ø Distributed switches(each node works as a switch) Ø High path Rack-Scale Network Topology Ø Distributed switches(each node works as a switch) Ø High path](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-19.jpg)
![Broadcasting-Assisted Rack Congestion Control Broadcasting overhead is low(around 1. 3%). Ø Broadcast flow information(e. Broadcasting-Assisted Rack Congestion Control Broadcasting overhead is low(around 1. 3%). Ø Broadcast flow information(e.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-20.jpg)
![Evaluation 21 Evaluation 21](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-21.jpg)
![Congestion Control for RDMA-enabled Datacenters Zhu, Yibo, et al. "Congestion Control for Large-Scale RDMA Congestion Control for RDMA-enabled Datacenters Zhu, Yibo, et al. "Congestion Control for Large-Scale RDMA](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-22.jpg)
![Congestion Spreading in Lossless Networks SE E P PAU E S PAU E A Congestion Spreading in Lossless Networks SE E P PAU E S PAU E A](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-23.jpg)
![Wireless Congestion Control Zaki, Yasir, et al. "Adaptive Congestion Control for Unpredictable Cellular Networks. Wireless Congestion Control Zaki, Yasir, et al. "Adaptive Congestion Control for Unpredictable Cellular Networks.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-24.jpg)
![What do Cellular Traffic Look Like? Burst Scheduling Competing Traffic 25 What do Cellular Traffic Look Like? Burst Scheduling Competing Traffic 25](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-25.jpg)
![What do Cellular Traffic Look Like? Channel Unpredictability 26 What do Cellular Traffic Look Like? Channel Unpredictability 26](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-26.jpg)
![Verus Protocol Epoch i+1 Sending window Wi+1 Wi Ø Epoch: a short period of Verus Protocol Epoch i+1 Sending window Wi+1 Wi Ø Epoch: a short period of](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-27.jpg)
![Verus Overview Delay Estimator: estimate delay in the future based on the changes of Verus Overview Delay Estimator: estimate delay in the future based on the changes of](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-28.jpg)
![Delay Estimation Epoch i-1 Epoch i Dmax, i = alpha x. Dmax, i-1 + Delay Estimation Epoch i-1 Epoch i Dmax, i = alpha x. Dmax, i-1 +](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-29.jpg)
![Window Update Ø Delay-Window Profile: updated based on historical data Ø Each epoch can Window Update Ø Delay-Window Profile: updated based on historical data Ø Each epoch can](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-30.jpg)
![Packet Scheduler Epoch i Sending window Wi Epoch i+1 Sending window Wi+1 Ø How Packet Scheduler Epoch i Sending window Wi Epoch i+1 Sending window Wi+1 Ø How](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-31.jpg)
![Loss Handling Epoch i Sending window Wi Epoch i+1 Multiplicative Decrease Wi+1 = M Loss Handling Epoch i Sending window Wi Epoch i+1 Multiplicative Decrease Wi+1 = M](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-32.jpg)
![Evaluation 33 Evaluation 33](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-33.jpg)
![Thanks! 34 Thanks! 34](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-34.jpg)
- Slides: 34
![Revisiting Transport Congestion Control Jian He UT Austin 1 Revisiting Transport Congestion Control Jian He UT Austin 1](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-1.jpg)
Revisiting Transport Congestion Control Jian He UT Austin 1
![Why is Congestion Control necessary Data Packets Congested Link ACK Ø Congested link vs Why is Congestion Control necessary? Data Packets Congested Link ACK Ø Congested link vs.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-2.jpg)
Why is Congestion Control necessary? Data Packets Congested Link ACK Ø Congested link vs. reliability: long queuing delay, packet l Ø But, can delay or packet loss always well explain congestio 2
![Can we distinguish congestion reasons Ø Congestion related signals packet loss duplicate ACKs Can we distinguish congestion reasons? Ø Congestion related signals: - packet loss: duplicate ACKs,](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-3.jpg)
Can we distinguish congestion reasons? Ø Congestion related signals: - packet loss: duplicate ACKs, retransmission timeout (TCP Reno, TCP Cubic) - round-trip delay: TCP packet RTT (TCP Vegas, FAST TCP, Compound TCP) - queue size: explicit congestion notification(ECN) (DCTCP) 3
![Existing TCP Variants TCP ThroughputLatency Tradeoff Exploration Remy SIGCOMM 13 Datacenter TCP Tail performanceTIMELY Existing TCP Variants TCP Throughput-Latency Tradeoff Exploration [Remy SIGCOMM’ 13] Datacenter TCP Tail performance[TIMELY](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-4.jpg)
Existing TCP Variants TCP Throughput-Latency Tradeoff Exploration [Remy SIGCOMM’ 13] Datacenter TCP Tail performance[TIMELY SIGCOMM’ 15], New Architectures[R 2 C 2 SIGCOMM’ 15] RDMA[DCQCN SIGCOMM’ 15] Persistently High Performance Large flows[PCC NSDI’ 15] Highly-variant network condition Cellular transport[Verus SIGCOMM’ 15, Sprout NSDI’ 13] Reducing Start-up Delay [Halfback Co. Next’ 15], [RC 3 NSDI’ 14] Performance interference for competing flows Application Heterogeneity[QJUMP NSDI’ 15] 4
![TCP Evolution ApplicationSpecific Performance Requirements Application Sensing Layer TCP Networking Sensing Layer IP Link TCP Evolution Application-Specific Performance Requirements Application Sensing Layer TCP Networking Sensing Layer IP Link](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-5.jpg)
TCP Evolution Application-Specific Performance Requirements Application Sensing Layer TCP Networking Sensing Layer IP Link Network Condition Hardware 5
![Optimizing Datacenter Transport Tail Performance Mittal Radhika et al TIMELY RTTbased congestion control for Optimizing Datacenter Transport Tail Performance Mittal, Radhika, et al. "TIMELY: RTT-based congestion control for](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-6.jpg)
Optimizing Datacenter Transport Tail Performance Mittal, Radhika, et al. "TIMELY: RTT-based congestion control for the datacenter In ACM SIGCOMM 2015. 6
![Why does tail performance matter Ø TCP Incast many servers reply the client Why does tail performance matter? … Ø TCP Incast: many servers reply the client](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-7.jpg)
Why does tail performance matter? … Ø TCP Incast: many servers reply the client simultaneously Ø All replies should meet their deadlines. Ø Datacenter transport must deliver high throughput(>>Gbps) and utilization with low delay(<<msec). 7
![Hardware Assisted RTT Measurement Why was RTT not widely used Ø RTTbased congestion control Hardware Assisted RTT Measurement Why was RTT not widely used? Ø RTT-based congestion control](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-8.jpg)
Hardware Assisted RTT Measurement Why was RTT not widely used? Ø RTT-based congestion control performed poorly at WANs. Ø Highly noisy RTT estimation(system kernel scheduling, etc. Ø Datacenter RTT measurement needs ms-level granularity. Ø Hardware timestamp and hardware acknowledgement can significantly remove noise. 8
![RTT As a Congestion Control Signal Multibit signal Singlebit signal Ø ECN can not RTT As a Congestion Control Signal Multi-bit signal Single-bit signal Ø ECN can not](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-9.jpg)
RTT As a Congestion Control Signal Multi-bit signal Single-bit signal Ø ECN can not reflect the extent of end-to-end latency inflated by network queuing, due to traffic priorities, multiple congested switches, etc. 9
![RTT Correlates with Queuing Delay 10 RTT Correlates with Queuing Delay 10](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-10.jpg)
RTT Correlates with Queuing Delay 10
![TIMELY Framework 11 TIMELY Framework 11](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-11.jpg)
TIMELY Framework 11
![RTT Measurement tsend Serialization Delay RTT tcompletion Propagation Queuing Delay ACK Turnaround Time RTT Measurement tsend Serialization Delay RTT tcompletion Propagation & Queuing Delay ACK Turnaround Time](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-12.jpg)
RTT Measurement tsend Serialization Delay RTT tcompletion Propagation & Queuing Delay ACK Turnaround Time Ø One RTT for one segment (NIC Offload) Ø Hardware ACKs make ACK turnaround time ignorable Ø RTT = Propagation + Queuing Delay = tcompletion – tsend – segment_size/NIC_line_rate 12
![Transmission Rate Control Message to be sent Segments RTT Estimation Rate Controller Insert delay Transmission Rate Control Message to be sent Segments RTT Estimation Rate Controller Insert delay](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-13.jpg)
Transmission Rate Control Message to be sent Segments RTT Estimation Rate Controller Insert delay between segments Transmission Queue Ø Target rate is determined by segment size and delay between segments 13
![Rate vs Window Ø Segment size as high as 64 KB Ø 32 us Rate vs. Window Ø Segment size as high as 64 KB. Ø (32 us](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-14.jpg)
Rate vs. Window Ø Segment size as high as 64 KB. Ø (32 us RTT x 10 Gbps) = 40 KB window size Ø 40 KB < 64 KB: Window makes no sense 14
![Rate Update 15 Rate Update 15](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-15.jpg)
Rate Update 15
![Evaluation 16 Evaluation 16](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-16.jpg)
Evaluation 16
![Datacenter Transport for Emerging Architectures Costa Paolo et al R 2 C 2 A Datacenter Transport for Emerging Architectures Costa, Paolo, et al. "R 2 C 2: A](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-17.jpg)
Datacenter Transport for Emerging Architectures Costa, Paolo, et al. "R 2 C 2: A Network Stack for Rack-scale Computers. " In ACM SIGCOMM 2015. 17
![RackScale Computing Ø Building Block for future datacenters Ø High BW low latency network Rack-Scale Computing Ø Building Block for future datacenters Ø High BW low latency network](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-18.jpg)
Rack-Scale Computing Ø Building Block for future datacenters Ø High BW low latency network Ø Direct-connected topology 18
![RackScale Network Topology Ø Distributed switcheseach node works as a switch Ø High path Rack-Scale Network Topology Ø Distributed switches(each node works as a switch) Ø High path](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-19.jpg)
Rack-Scale Network Topology Ø Distributed switches(each node works as a switch) Ø High path diversities 3 D Torus Fat-tree Topology 19
![BroadcastingAssisted Rack Congestion Control Broadcasting overhead is lowaround 1 3 Ø Broadcast flow informatione Broadcasting-Assisted Rack Congestion Control Broadcasting overhead is low(around 1. 3%). Ø Broadcast flow information(e.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-20.jpg)
Broadcasting-Assisted Rack Congestion Control Broadcasting overhead is low(around 1. 3%). Ø Broadcast flow information(e. g. , start time, finish time) Ø Each node has a global view of the network Ø Locally optimize flow rate with the global view 20
![Evaluation 21 Evaluation 21](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-21.jpg)
Evaluation 21
![Congestion Control for RDMAenabled Datacenters Zhu Yibo et al Congestion Control for LargeScale RDMA Congestion Control for RDMA-enabled Datacenters Zhu, Yibo, et al. "Congestion Control for Large-Scale RDMA](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-22.jpg)
Congestion Control for RDMA-enabled Datacenters Zhu, Yibo, et al. "Congestion Control for Large-Scale RDMA Deployments. ” In ACM SIGCOMM, 2015. 22
![Congestion Spreading in Lossless Networks SE E P PAU E S PAU E A Congestion Spreading in Lossless Networks SE E P PAU E S PAU E A](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-23.jpg)
Congestion Spreading in Lossless Networks SE E P PAU E S PAU E A U S US PAUSE PA PAUSE E P US A U SE PA P A U SE SE PAU Ø Port-based congestion control incurs congestion spreading Ø DCQCN: incorporating explicit congestion notification to support flow-based congestion control 23
![Wireless Congestion Control Zaki Yasir et al Adaptive Congestion Control for Unpredictable Cellular Networks Wireless Congestion Control Zaki, Yasir, et al. "Adaptive Congestion Control for Unpredictable Cellular Networks.](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-24.jpg)
Wireless Congestion Control Zaki, Yasir, et al. "Adaptive Congestion Control for Unpredictable Cellular Networks. “ In SIGCOMM 2015. 24
![What do Cellular Traffic Look Like Burst Scheduling Competing Traffic 25 What do Cellular Traffic Look Like? Burst Scheduling Competing Traffic 25](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-25.jpg)
What do Cellular Traffic Look Like? Burst Scheduling Competing Traffic 25
![What do Cellular Traffic Look Like Channel Unpredictability 26 What do Cellular Traffic Look Like? Channel Unpredictability 26](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-26.jpg)
What do Cellular Traffic Look Like? Channel Unpredictability 26
![Verus Protocol Epoch i1 Sending window Wi1 Wi Ø Epoch a short period of Verus Protocol Epoch i+1 Sending window Wi+1 Wi Ø Epoch: a short period of](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-27.jpg)
Verus Protocol Epoch i+1 Sending window Wi+1 Wi Ø Epoch: a short period of time (e. g. , 5 ms) Ø Sending window is updated at each epoch. Ø Sending window represents the number packets in flight. 27
![Verus Overview Delay Estimator estimate delay in the future based on the changes of Verus Overview Delay Estimator: estimate delay in the future based on the changes of](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-28.jpg)
Verus Overview Delay Estimator: estimate delay in the future based on the changes of delay Delay Profiler: record the relationship of delay-sending window Go to next epoch Window Estimator: estimate the sending window for the next epoch Packet Scheduler: calculate the number packets to be sent in the next epoch 28
![Delay Estimation Epoch i1 Epoch i Dmax i alpha x Dmax i1 Delay Estimation Epoch i-1 Epoch i Dmax, i = alpha x. Dmax, i-1 +](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-29.jpg)
Delay Estimation Epoch i-1 Epoch i Dmax, i = alpha x. Dmax, i-1 + (1 -alpha) x. Dmax, i ∆Di = Dmax, i -Dmax, i-1 ∆Di<=0 Estimated Delay Dest, i • ∆Di>0 • Dest, i+1 • Time 29
![Window Update Ø DelayWindow Profile updated based on historical data Ø Each epoch can Window Update Ø Delay-Window Profile: updated based on historical data Ø Each epoch can](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-30.jpg)
Window Update Ø Delay-Window Profile: updated based on historical data Ø Each epoch can contribute many points to the profile. Ø Profile is initialized using data in the slow-start phase. 30
![Packet Scheduler Epoch i Sending window Wi Epoch i1 Sending window Wi1 Ø How Packet Scheduler Epoch i Sending window Wi Epoch i+1 Sending window Wi+1 Ø How](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-31.jpg)
Packet Scheduler Epoch i Sending window Wi Epoch i+1 Sending window Wi+1 Ø How many packets to be sent in current epoch? Si+1 = max[0, (Wi+1 + ((2 -n)/(n-1))*Wi)] n is the number of epochs over the current estimated RTT 31
![Loss Handling Epoch i Sending window Wi Epoch i1 Multiplicative Decrease Wi1 M Loss Handling Epoch i Sending window Wi Epoch i+1 Multiplicative Decrease Wi+1 = M](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-32.jpg)
Loss Handling Epoch i Sending window Wi Epoch i+1 Multiplicative Decrease Wi+1 = M * Wi Ø Stop updating delay profile during the loss recovery phase 32
![Evaluation 33 Evaluation 33](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-33.jpg)
Evaluation 33
![Thanks 34 Thanks! 34](https://slidetodoc.com/presentation_image_h/c09b601e7209ba586b6f9c13993e4e23/image-34.jpg)
Thanks! 34
Pincuegula
Jian-yun nie
Jian-jiun ding
Jian-jiun ding
Activities in jian
Jian sheng wang
Jian sun
General principles of congestion control
In2140
Tcp congestion control
Traffic throttling and load shedding
Principles of congestion control
Congestion control principles
Congestion control in virtual circuit
Udp congestion control
New reno tcp
Principles of congestion control
Congestion control in network layer
Choke packets in computer networks
Segment header
General principles of congestion control
Traffic control austin
Primary and secondary transport
Primary vs secondary active transport
Now answer the following questions
Active vs passive transport venn diagram
Endocytosis vs exocytosis
Primary active transport vs secondary active transport
Bioflix activity membrane transport active transport
Active and passive transport
Selectively permeable definition biology
Network congestion causes
Cause and effect introduction
Conclusion of traffic congestion
Cvc lung microscopy