Strategies for Caching Information on Distributed Systems

由于负担得起的高速互联网技术、高性能计算机和外围设备以及万维网的显著进步，数字信息已成为我们日常生活中不可分割的一部分。Web的爆炸性增长对网络资源和Web服务器提出了巨大的需求，导致远程Web服务请求的长时间延迟。本项目将开发一个集成的研究和教育计划，研究各种有效和高效的分布式系统信息缓存策略，如Web缓存，以减少服务延迟和广域网络带宽消耗。研究工作将集中在：(I)深入了解Web使用的重要和未探索的特征；(Ii)开发这些特征的准确性能模型，以衡量现有和新开发的缓存策略的有效性；(Iii)设计新的负载平衡技术，最大限度地利用包括缓存服务器和客户端在内的分布式系统的聚合能力；以及(Iv)开发可扩展的视频缓存技术。这些技术的性能将通过模拟和实验进行测量。这些研究工作将对网络的发展产生重大影响。该项目的教育部分围绕着各种策略，以激励和留住潜在的学生参加强调高质量研究的研究生项目。通过研究和教育工作的健康协同，PI将通过教学和学生指导影响学生，为研究社区做出贡献，并影响普通公众。

Digital information has become an inseparable part of our daily life due to significant advances in affordable high-speed internetworking technologies, high performance computer and peripherals, and the World-Wide Web. The explosive growth of the Web has imposed a heavy demand on networking resources and Web servers, resulting in long delays when requesting remote Web services.This project will develop an integrated research and education program that investigates various effective and efficient strategies for caching information on distributed system such as Web caching to reduce service delays and wide-area-network bandwidth consumption. The research work will focus on (i) gaining a deep understanding of important and unexplored characteristics of Web usage; (ii) developing accurate performance models of these characteristics to measure the effectiveness of existing and newly developed caching strategies; (iii) designing novel load-balancing techniques that best utilize the aggregated capabilities of distributed systems including cache servers and client machines; and (iv) developing scalable technologies for video caching. Performance of these techniques will be measured via simulation and experimentation. The research efforts will have a significant impact on the evolution of the Web.The education part of the program is centered around various strategies to motivate and retain potential students for the graduate program that emphasizes quality research. Through a healthy synergy of both research and education efforts, the PI will influence students through teaching and student mentoring, contribute to the research community, and impact the general public.