|
Abstract : |
Network traffic locality is a special case of the locality phenomenon commonly seen in computer systems. The source and destination addresses seen in network packet traffic are observed to be non-uniformly distributed, both in time and space. Several recent research efforts have identified the presence of network traffic locality, and suggested ways to exploit the locality property in the design and operation of computer networks. While the intuitive notion of network traffic locality is fairly well understood, there is still a distinct lack of formal definitions or measures of network locality in the literature. In this paper, we propose four new measures to characterize network traffic locality: persistence, concentration, address re-use, and reference density. These measures are motivated, in part, by locality characteristics identified in other domains (e.g., memory reference studies of program behaviour, and file referencing characteristics), but reflect properties important to network applications. Definitions are given for each measure, and measurement data collected from a local area network environment are used to quantify each measure. Finally, to demonstrate a potential application for the network traffic locality property, we present a brief analysis of how this property might be used to advantage in the design of a network interconnection device., |
