Erasure Coding and Congestion Control for Interactive Real-Time Communication

The use of real-time applications over the Internet is a challenging problem that the QoS epoch attempted to solve by proposing the DiffServ architecture. Today, the only existing service provided by the Internet is still best-effort. As a result, multimedia applications often perform on top of a transport layer that provides a variable sending rate. In an obvious manner, this variable sending rate is an issue for these applications with strong delay constraint. In a real-time context where retransmission can not be used to ensure reliability, video quality suffers from any packet losses. To illustrate this problem and motivate why we want to bring out a certain class of erasure coding scheme inside a multimedia congestion control protocol such as TFRC [1], we propose this simple simulation scenario: we encode a CIF ’Foreman’ sequence of 300 frames with two values of Quantization Parameter (QP). For a Group of Picture (GOP) size of 30 frames, QP 28 and 29 result in H.264/AVC videos of 531.13 kb/s (Peak Signal to Noise Ratio (PSNR) of 36.9 dB) and 441.37 kb/s (PSNR of 36.2 dBs), respectively. A video with QP of 28 generates 16.9 % of redundancy bit rate compared to the one of 29. We drive the simulations with encoded video of QP 29 protected by an Erasure Coding (Tetrys mechanism [11]) with redundancy of 12.5% and the one of 28 without protection. We randomly generate a loss rate of 1%. In Fig. 1, we observe a strong quality degradation of the video without protection (blue line compared to purple line) while Erasure Coding recovers all losses within one way End-to-End delay constraint of 150 ms (red and green lines are identical). This implies that a slightly higher quality without protection suffers from any losses while a slightly lower quality video with protection can sustain up to a certain loss rate (3% in these simulations) depending on its redundancy. As a matter of fact, erasure-coding seems mandatory to enable a fair real-time video quality.