An Integrated Source Transcoding and Congestion Control Paradigm for Video Streaming in the Internet Proposed by R. Puri, K.W. Lee, K. Ramchandran and V. Bharghavan Presented by Felix

Agenda  Introduction  LIMD/H Congestion Control  MD-FEC Transcoder  Performance Summary  Discussion

Introduction  More and more video traffic are using best-effort and unreliable channel service (eg. UDP) because of its low delay feature  Explicit TCP-friendly Congestion Control policy is required to achieve fairness with other TCP flows

Introduction  Two questions: How to control the transmission rate? How to control the video source to adapt to the transmission rate?

LIMD/H Congestion Control  LIMD (Linear Increase Multiplicative Decrease) r : sending rate f : fraction of packet loss in the last period If f = 0, r  r + a If f > 0, r  r * (1-b) Typically, (a, b) = (1, 0.5)

LIMD/H Congestion Control  Problems of traditional LIMD React identically and aggressively to any kind of packet loss, both congestion induced and non-congestion induced. Even if the channel bandwidth is invariant, the sending rate fluctuates greatly

LIMD/H Congestion Control  LIMD/H (LIMD with History) h : a history factor If f = 0, r  r + a, and h  1 If f > 0, r  r *(1 – b ’ *h), and h  2h b ’ should be small to reduce the variation of sending rate

LIMD/H Congestion Control

MD-FEC Transcoder  Layered or Multi-Resolution (MR) source coding is a common coding method to provide quality/bit-rate scalability Multiresolution (MR) Source Coder.... Base Layer Enhancement Layer 1 Enhancement Layer 2 Enhancement Layer N Raw Video Stream

MD-FEC Transcoder  Problem of MR coding Different quality layers have different importance  Prioritized Eg. The client receiving packets of layers [0, 1, 2, 4] will only get quality as just receiving [0, 1, 2] However, the network treats every packet, no matter which layer it belongs to, identically  Non-prioritized Loss of lower layers ’ packets makes some other successfully transmitted higher layers ’ packets useless. Low robustness in a lossy channel

MD-FEC Transcoder  MD-FEC Transform a prioritized MR bit-stream to non- prioritized MD (Multi-Description) stream with additional redundancy (using FEC)

MD-FEC Transcoder  Step 1: Partition a MR bit-stream to N layers and split the ith layer into i equal parts 23 … i … 1 N R0R0 R i-1 R2R2 R i-2 R1R1 R N-1 … R N … i-1i …

MD-FEC Transcoder  Step 2: Adopt (N, i, N-i+1) Reed-Solomon code to the ith layer and form N packets as follows: 12 … i … N FEC2 … i … N … i … N … … N … … N Packet 1 Packet 2 Packet i Packet i+1 Packet N

MD-FEC Transcoder  Optimization on R i, i=0…N Notations:  q i (N) : probability that i+1 out of N packets are delivered to the destination  D(r) : Distortion function of rate r Rate Distortion R0R0 R1R1 … RiRi R N-1 …

MD-FEC Transcoder Cont ’ Problem statement:  Minimize the expected distortion ED:  E is the distortion encountered when the source is represented by zero bits

MD-FEC Transcoder Cont ’ The total rate R t equals: Thus constraints to the optimization problem are:

Performance Summary  Simulations have been done to show the variation of sending rate and PSNR upon variation in network capacity and random losses  MD-FEC can maintain a smooth PSNR  LIMD/H can adapt to the variation of network bandwidth quickly while reducing the rate fluctuation induced by random losses and channel probing

Discussion  The MD-FEC is a novel technique to add robustness to many layer coding schemes  LIMD/H congestion control mechanism can provide low variation in transmission rate while guaranteeing inter-traffic fairness  However, the delay caused by explicit end-to-end feedback and FEC operations may affect the performance of the system