Abstract

Background Researchers widely use networked cybersecurity testbed facilities, such as emulab-based testbeds, to conduct scientific experiments. Experimentation tools provide a high level logical view of the testbed resources that abstract away the heterogeneity of the underlying hardware and software. These abstractions are highly effective in enabling cybersecurity experimentation. However, currently there are no tools to characterize the collective properties of the assigned resources, and to understand their impact on the results. Aim In order to conduct a scientifically valid experiment, there is a need to understand the properties of the assigned testbed resources which form the " scientific instrument" for the experiment. In this paper, we propose a calibration methodology to allow experimenters to characterize, interpret, and adjust the experimental results by accounting for the impact of the underlying resources. Method We developed a calibration tool to characte...
Background

Aim

Method

Date

2014

Authors

Vineet Ghatge