On modeling and evaluating multicomputer transcoding architectures for live-video streams

Rodrigo Fernandes de Mello, Rudinei Goularte, Evgueni Dodonov, Laurence Tianruo Yang, Jong Hyuk Park, Tai-Hoon Kim

Abstract: The pervasive and ubiquitous computing has motivated researches on multimedia adaptation which aims at matching the video quality to the user needs and device restrictions. This technique has a high computational cost which needs to be studied and estimated when designing architectures and applications. This paper presents an analytical model to quantify these video transcoding costs in a hardware independent way. The model was used to analyze the impact of transcoding delays in end-to-end live-video transmissions over LANs, MANs and WANs. Experiments confirm that the proposed model helps to define the best transcoding architecture for different scenarios.