Improving video quality in congested networks through deferred discard

Video already constitutes more than half of Internet traffic, and is expected to expand to 86% of Internet traffic by 2016. The dominance of video traffic justifies router features that optimise performance when video loads are high, in particular features that can deal with the congestion that video can create. In this paper, we exploit a characteristic of video, namely that video streams consist of recurring patterns of packets of varying importance, to create a new Deferred Discard (DD) queue management scheme. DD keeps important video packets when queue management would otherwise have discarded them, and substitutes their discard with the discard of later less important packets. We simulated different queue management schemes with a set of three test videos and measured their performance in terms of the average picture signal to noise ratio (PSNR). DiffServ improved the video PSNR by 2.1dB compared to best effort, while DD further improved the video PSNR (of DiffServ) by 3.9dB, i.e. DD improves video quality over DS more than DS improves video quality over BE.