Efficient Web Browsing for Mobile Clients using HTTP Compression

Abstract

Efficient web browsing on mobile computers presents a unique challenge. These machines are different from other classes of client computers since they have relatively low-bandwidth connections and they are battery-powered and therefore limited by their energy consumption. However, they tend to interact with the same servers for the delivery of web content. This project investigates optimizing the final critical link between a mobile client and a stationary base station by compressing HTTP request and response messages. Using a split proxy design, compression of individual request messages reduces bandwidth by 26% to 34% across a variety of benchmark traces, and applying compression to response messages yields savings of 59% to 82% of the compressible data. Higher compression rates are achieved by using streaming compression algorithms to compress the streams of request and response messages. In this case, the bandwidth for requests sees an order of magnitude improvement, and the response stream obtains additional savings of 7% to 25% on top of the savings achieved with per-response compression.