A Framework for pervasive web content delivery

Abstract

The integration of the Internet and the wireless network causes Web clients more heterogeneous than ever. That challenge is further complicated by increased multimedia data traffic, compellingly personalized content, and demand for efficient Web content delivery. Pervasive services in Web content delivery are needed. This thesis addresses the challenges by suggesting the basic component a??- a fine-grained, scalable data model. Based on the data model, a novel adaptation called modulation was proposed. Modulation is characterized as fast, exclusive, and reversible. In the evaluation, modulation (using JPEG 2000 as an illustration) evidently outperforms transcoding (using JPEG) -a?? the oft-cited content adaptation. To replicate modulation's benefits to Web content delivery, the existing framework needs modifications. The proposed framework emphasizes collaboration between the origin server and the proxy, and exploits high data reuse. Evaluation shows that the new framework yields improved client perceived latency, conserved Internet bandwidth, and reduced server's load.