ns-2 Distributed Clients Emulation: Accuracy and Scalability

Abstract

ABSTRACTns-2 is a well known network simulator, recently extended with improvements to its emulation facility. Real-time constraints and the boundary between real-world and simulated entities impose scalability and accuracy limitations, and distort the simulated network as perceived by the involved realworld applications. This paper presents results from a performance evaluation of the ns-2 emulation facility. Conducting emulation experiments of differing magnitudes, and under varying emulation environment set-ups, we unveil central types of scalability limitations and obtainable accuracy. We find throughput limits using high and low end computers, and a significant throughput decrease when increasing the number of involved real-world applications. We furthermore show how end-to-end delay increases both with traffic load and an increasing number of involved real-world applications. Moreover, during these conditions, we find that the system treats these applications increasingly unfair by distributing total throughput unevenly between them, and by imposing different amounts of end-to-end delay.

Accuracy and scalability of ns-2’s distributed emulation extension

“…We have N pairs of TG instances (sources and destinations), communicating across separate paths (EmuPaths) through the SimNet, each identified by a unique index. The TAP devices at both ends of the EmuPaths are identified by the combination of an EmuPath index and the RWHost index, both integers in the range [1,255].…”

“…The main contributions of this paper are: (1) to aid researchers in the decision on whether or not the DCE provides the accuracy and scalability required for their own experiments, (2) to pinpoint the scalability limits of the DCE so as to help researchers avoid the pitfall of designing experiments too large to obtain reliable results, and (3) to show general considerations that are important during the design and implementation of distributed emulation systems similar to the DCE. From the insight gained during the analysis of our results, we also propose a set of modifications to the system believed to increase performance.…”