On the performance of caching in information-centric networks

Abstract

Whether the architecture of the future Internet is an incremental evolution of IP or a clean slate design like CCN, it is clear that caching will be increasingly used to reduce the amount of content traffic the network has to carry. It is important to understand the performance of caching in order to quantify its impact on traffic flows and to draw correct conclusions regarding the optimal network structure. We discuss recent work on modelling cache performance, identifying the crucial importance of how user requests are distributed over the vast catalogue of available content and stressing the difficulty in estimating this popularity distribution. Lastly, we consider the memory-bandwidth tradeoff that determines the potential cost advantage of investing in storage capacity rather than in the infrastructure that would otherwise be necessary to transport content across the network. Our evaluations suggest large caches at the network edge are preferable to equipping routers with limited capacity content stores, as envisaged in some proposed architectures.

James Roberts

James Roberts very recently joined the French research institute IRT-SystemX to work on a project on Cloud computing and network architecture. He was previously with Inria from September 2009 after spending more than thirty years with France Telecom research labs. He received a degree in Mathematics from the University of Surrey in 1970 and a doctorate in computer science in 1987 from the University of Paris VI. His research is centered on the performance evaluation and design of traffic controls for communication networks. In a long career, he has published around 100 papers, chaired several program committees and been associate editor for a number of journals. He gave the Keynote at Infocom 2013. He is a Fellow of the Société des électriciens et électroniciens (SEE) and recipient of the Arne Jensen lifetime achievement award from the International Teletraffic Congress (ITC).