Abstract

Image transcoding proxies are used to improve Web browsing over low bandwidth networks by adapting content-rich web images to bandwidth-constrained clients. Such transcoding proxies dynamically analyze, manipulate and transcode images (e.g. quality reduction, down sampling) on the fly enabling significant reductions in download times over low bandwidth links. However, transcoding proxies have scalability problems if the objective policy that decides whether to transcode an image does not take client load (e.g. number of concurrent clients) into consideration. We show that seemingly intuitive policies that decide solely based on whether transcoding yields savings in transmission time fail to scale. Transcoding performance can be improved by a factor of two in terms of average Web page latency as perceived by a client, by taking overall client load into consideration and by properly scheduling transcoding operations on a single CPU. We also show that an Earliest Deadline First Based (EDF) scheduling policy further improves transcoding performance.

Citations