40
Realtime Multimedia Streaming over Internet Pengjun Pei Dazhen Pan CSE 620 Fall,2001

Realtime Multimedia Streaming over Internet

  • Upload
    ivrit

  • View
    50

  • Download
    0

Embed Size (px)

DESCRIPTION

Realtime Multimedia Streaming over Internet. Pengjun Pei Dazhen Pan CSE 620 Fall,2001. Overview. Wide Application Video-conference Internet telephony Streaming audio/video players Challenges:Internet is best-effort network Packet loss Bandwidth variation Packet delay variation. - PowerPoint PPT Presentation

Citation preview

Page 1: Realtime Multimedia Streaming over Internet

Realtime Multimedia Streaming over Internet

Pengjun PeiDazhen Pan

CSE 620 Fall,2001

Page 2: Realtime Multimedia Streaming over Internet

Overview

Wide Application Video-conference Internet telephony Streaming audio/video players

Challenges:Internet is best-effort network Packet loss Bandwidth variation Packet delay variation

Page 3: Realtime Multimedia Streaming over Internet

System Architecture

Page 4: Realtime Multimedia Streaming over Internet

Content

Video CompressionCongestion ControlError Control

Page 5: Realtime Multimedia Streaming over Internet

Video Compression

Various requirement: Bandwidth Delay Loss VCR like function Decoding complexity

Intra-frame redundancy & inter-frame redundancyNon-scalable coding vs Scalable coding

Page 6: Realtime Multimedia Streaming over Internet

Inter-frame redundancy

MPEG-2:

I frame: intra-picture P frame: predicted pictureB frame: bi-directional predicted pictureMPEG frame dependencies in an MPEG bit

stream

Page 7: Realtime Multimedia Streaming over Internet

Scalable Coding

FGS: fine granularity scalability(proposed to MPEG-4):

Bitplanes of enhancement DCT coeffients

Page 8: Realtime Multimedia Streaming over Internet

Content

Video CompressionCongestion ControlError Control

Page 9: Realtime Multimedia Streaming over Internet

Congestion Control

Requirements for multimedia streaming Relatively constant rate Low latency for packet delivery Small latency variance Timely delivery is more important than

complete reliability

Rate controlRate shaping

Page 10: Realtime Multimedia Streaming over Internet

TCP/UDP?

TCP Retransmission mechanism

intolerable delays Multiplicative decrease in case of

congestionsharp variation in visual effect

UDP Unfair to responsive TCP flows Congestion collapse

Page 11: Realtime Multimedia Streaming over Internet

Categories of Rate Control

Source-based rate controlSource adjusts sending rateFeedback employedCan be applied to both unicast and

multicast

Receiver-based rate controlReceiver joins layer/channel Used in multicasting scalable video

Hybrid rate control

Page 12: Realtime Multimedia Streaming over Internet

TCP Friendly Flows

A flow is TCP-friendly if its arrival rate does not exceed the bandwidth of a conformant TCP connection in the same circumstances.

Page 13: Realtime Multimedia Streaming over Internet

TCP throughput model

λ: throughput of a TCP connection

MTU: maximum transit unitRTT: Round Trip Timep: packet loss ratio

pRTT

MTU

*

*325.1

Page 14: Realtime Multimedia Streaming over Internet

RAP(Rate Adaptation Protocol)

Proposed by R. Rejaie 1998End-to-end architecture

Page 15: Realtime Multimedia Streaming over Internet

RAP

Decision function If no congestion is detected, periodically increase the

transmission rate If congestion is detected, immediately decrease the

transmission rate

Increase/Decrease algorithm: AIMDDecision frequency

Smoothed version of one RTT: most recent value of SRTT

Page 16: Realtime Multimedia Streaming over Internet

RAPDecision function

Mechanisms to detect loss: Timeout

SRTTi = 7/8 * SRTTi + 1/8 * SampleRTTTimeout=μ*SRTT+δ*VarRTTUse transmission history ’coz it isn’t ack-clockedBefore sending a new packet, source traverses through the transmission history and detects all timeout losses: WHILE (DepartTimei+Timeout>=CurrTime)

IF(Flagi!=Acked) THENSeqi is lost

Detect a burst of loss at once

Page 17: Realtime Multimedia Streaming over Internet

RAP Decision function(Continued)

Gaps in sequence number(ACK-based)

ACK Packet:Acurr:packet being acknowledged

N: the last packet before Acurr that was still missing

Alast:the last packet before N that was received Timeout mechanism as a backup for critical

scenarios such as when a burst of packets is lost

Page 18: Realtime Multimedia Streaming over Internet

AIMD in RAP

No-packet loss: Si = Si + α (step height) Si = PacketSize/IPGi IPGi+1 = IPGi*C/( IPGi + C ) α = Si+1 – Si = PacketSize/C

Upon packet loss: Si+1 = β*Si IPGi+1 = IPGi/β β = 0.5

IPG:inter-packet-gap

Page 19: Realtime Multimedia Streaming over Internet

RAP Decision Frequency

Adjust IPG once every round-trip time using most recent value of SRTTRight value of C: C must be adjusted so that in a steady state,

the number of packets transmitted per step is increased by 1.

If IPG is updated once every T seconds and we choose C = T/k, the # of packets sent during each step is increased by k every step.

RAP use k=1 to emulate the TCP window adjust

Page 20: Realtime Multimedia Streaming over Internet

RAP Fine-grain rate adaptation

Motivation:Make RAP more stable and responsive to transient

congestion while still performing the AIMD algorithm at a coarser granularity

Fine-grain feedback:Feedbacki=FRTTi/XRTTiFRTTi,XRTTi: short/long term exponential moving

average of RTT samples at the ith adjusting pointRTTi+1 = (1 – K)RTTi+K*SampleRTT(KXRTT=0.01 KFRTT=0.9)

Fine-grain adjustmentIPGi’ = IPGi * Feedbacki

Page 21: Realtime Multimedia Streaming over Internet

Simulation Result RAP

Page 22: Realtime Multimedia Streaming over Internet

Simulation Result(FG-RAP)

Page 23: Realtime Multimedia Streaming over Internet

Binomial Algo

Proposed by D. Bansal 2000

I: Increase in window as a result of receipt of one window of ACK in a RTTD:Decrease in window on detection of a loss by the senderWt: window size at time t

10:

0;/:

ltttt

kttRt

D

I

Page 24: Realtime Multimedia Streaming over Internet

Properties of Binomial Algo

Any l < 1 has a decrease that is in general less than a multiplicative decreaseTCP Friendly if and only ifk + l = 1 and l <= 1 for suitable α and β.Converge to fairness as long ask > =0, l >= 0, k + l > 0

10:

0;/:

ltttt

kttRt

D

I

Page 25: Realtime Multimedia Streaming over Internet

Ratio of throughput AIMD/Binomial

x:value of ky:TCP throughput/Binomial throuput

Page 26: Realtime Multimedia Streaming over Internet

SQRT(k = l = 0.5)

Page 27: Realtime Multimedia Streaming over Internet

SQRT vs AIMD

SQRT has less oscillatory bandwidth probing

Page 28: Realtime Multimedia Streaming over Internet

Categories of Rate Control

Source-based rate controlSource adjusts sending rateFeedback employedCan be applied to both unicast and

multicast

Receiver-based rate controlReceiver joins layer/channel Used in multicasting scalable video

Hybrid rate control

Page 29: Realtime Multimedia Streaming over Internet

Source-based Rate Control for Multicast

Unicast video distribution using multiple point-point connectionMulticast video distribution using point-to-multipoint transmission

Page 30: Realtime Multimedia Streaming over Internet

Single-Channel Multicast

IVS(INRIA Video-conference System): Single-channel multicast Probe-base,use AIMD Each receiver determine the network

status Source solicits network status info

through probabilistic polling to avoid feedback implosion

Compare the fraction of congested receiver with threshold

Page 31: Realtime Multimedia Streaming over Internet

Multiple-channel multicast

Differentiated service to receivers because each receiver can individually negotiate service parameters with the recourseBandwidth inefficiency

Page 32: Realtime Multimedia Streaming over Internet

Categories of Rate Control

Source-based rate controlSource adjusts sending rateFeedback employedCan be applied to both unicast and

multicast

Receiver-based rate controlReceiver joins layer/channel Used in multicasting scalable video

Hybrid rate control

Page 33: Realtime Multimedia Streaming over Internet

Receiver-based Rate Control

Typically applied to layered multicast video Source-based works reasonably well for

unicast Receiver-based targeted at solving

heterogeneity problem in the multicast case

Probe-based: No congestion, receiver probes for available

bandwidth by joining layer/channel When congested,receiver drops a layer

Page 34: Realtime Multimedia Streaming over Internet

Receiver-based Rate Control

Model-based: Based on throughput model of TCP γi:transmission rate of Layer I, L current highest

layer Starts with subscribing base layer(Layer 0), set

L=0. Obtain MTU,RTT, p for a given period, calculate

throughput λ . If λ < γ0 drop base layer and stop receiving

video Else determine L’ ,the largest integer such that

'

0

L

i i

Page 35: Realtime Multimedia Streaming over Internet

Categories of Rate Control

Source-based rate controlSource adjusts sending rateFeedback employedCan be applied to both unicast and

multicast

Receiver-based rate controlReceiver joins layer/channel Used in multicasting scalable video

Hybrid rate control

Page 36: Realtime Multimedia Streaming over Internet

Hybrid Rate Control

Targeted at multicast videoApplicable to both layered video and non-layered videoMultiple channels, sender dynamically adjusts the rate for each channelDSG(Destination Set Grouping) Multiple streams:same video info with different

rate and quality,each sent to an IP multicast group Receiver chooses a multicast group to join Source uses feedback to adjust rate for each

stream

Page 37: Realtime Multimedia Streaming over Internet

Rate Shaping

Adapt the rate of compressed video bit-streams to the target rate constraint

Page 38: Realtime Multimedia Streaming over Internet

Types of rate filter

Codec Filter: perform transcoding between different schemesFrame-dropping filter:distinguish frame types and drop frames according to importanceLayer-dropping filter:distinguish layers and drop frames according to importanceFrequency filter:discard DCT coefficient of high frequencies

Page 39: Realtime Multimedia Streaming over Internet

Conclusion

Binomial rate control causes less oscillation to multimedia stream

Current research separates rate control and rate shaping

Page 40: Realtime Multimedia Streaming over Internet

Main ReferenceDeepak Bansal and Hari Balakrishnan ,Binomial Congestion ControlProc. IEEE INFOCOM Conf., Anchorage, AK, April 2001. S.Floyd, M. Handley,J.Padhye, and J. Widmer, Equation-Based Congestion Control for Unicast Applications, Proc. ACM SIGCOMM’00, pages43-54, Stockholm,Sweden,September 2000International Organization for Standardization. Overview of the MPEG-4 Standard,December 1999Sally Floyd,Kevin Fall, Promoting the Use of End-to-End Congestion Control in the Internet IEEE/ACM Transaction on Networking,May 1999Dapeng Wu,Yiwei Thomas Hou,etc, Streaming Video over the Internet: Approaches and Directions IEEE Transaction on Circuits and System for Video Technology, Vol11,No1,February 2001Dapeng Wu, Yiwei Thomas Hou,etc, Transorting real-time video over the Internet:challenges and approaches, Proceedings of the IEEE, vl.88,no. 12, Dec.2000R.Padhye,J.Kurose,D.Towsley,and R.Koodi. A Model-based TCP-Friendly Rate Control Protocol. In Proc. IEEE NOSSDAV’99,Basking Ridge,New Jersey,June 1999R.Rejaie,M.Handely,and D.Estrin. RAP:An End-to-end Rate-based Congestion Control Mechanism for Realtime Streams in the Internet. In Proc. IEEE Infocom’99,New York,NY,March 1999