MONROE-Nettest: A Configurable Tool for Dissecting Speed Measurements in Mobile Broadband Networks (1710.07805v2)

Abstract: As the demand for mobile connectivity continues to grow, there is a strong need to evaluate the performance of Mobile Broadband (MBB) networks. In the last years, mobile "speed", quantified most commonly by data rate, gained popularity as the widely accepted metric to describe their performance. However, there is a lack of consensus on how mobile speed should be measured. In this paper, we design and implement MONROE-Nettest to dissect mobile speed measurements, and investigate the effect of different factors on speed measurements in the complex mobile ecosystem. MONROE-Nettest is built as an Experiment as a Service (EaaS) on top of the MONROE platform, an open dedicated platform for experimentation in operational MBB networks. Using MONROE-Nettest, we conduct a large scale measurement campaign and quantify the effects of measurement duration, number of TCP flows, and server location on measured downlink data rate in 6 operational MBB networks in Europe. Our results indicate that differences in parameter configuration can significantly affect the measurement results. We provide the complete MONROE-Nettest toolset as open source and our measurements as open data.