Load Characterization and Measurement

Abstract

Timely detection of changes in traffic is critical for initiating appropriate traffic engineering mechanisms. Accurate measurement of traffic is an essential step towards change detection and traffic engineering. However, precise traffic measurement involves inspecting every packet traversing a link, resulting in significant overhead, particularly on routers with high speed links. Sampling techniques for traffic load estimation are proposed as a way to limit the measurement overhead. Since the efficacy of change detection depends on the accuracy of traffic load estimation, it is necessary to control error in estimation due to sampling. In this paper, we address the problem of bounding sampling error within a pre-specified tolerance level. We derive a relationship between the number of packet samples, the accuracy of load estimation and the squared coefficient of variation of packet size distribution. Based on this relationship, we propose an adaptive random sampling technique that determines the minimum sampling probability adaptively according to traffic dynamics. Using real network traffic traces, we show that the proposed adaptive random sampling technique indeed produces the desired accuracy, while also yielding significant reduction in the amount of traffic samples, yet simple to implement.We also investigate the impact of sampling errors on the performance of load change detection.