TU Wien, Institute of Telecommunications

A1 Telekom Austria

The capacity of the connecting network links mainly defines the quality of service for an end user. There are several parameters acting network communication links, an important one being the available end-to-end bandwidth. Several different approaches for measuring the network link capacity exist in industry, standardisation and the scientific community. However, their implementations are resource demanding and require a long measurement time. Both properties do not deal well with the challenges found in mobile broadband networks, especially when targeting a distributed measurement system. Bandwidth measurements in cellular wireless networks, with traffic reactive nature, nomadic end users, tariff data rate limitations and highly dynamic shared resources, is a challenging task.

The goal of FaLiCaB was to integrate a packet burst based solution, refined by parameter estimation into the MONROE platform. The MONROE platform offered a unique opportunity for the FaLiCaB consortium to validate the new method in nomadic environments. The active performance measurements, including estimation of available bandwidth, should be carried out by any user with the actual device they use to connect to modern mobile networks, e.g., smartphones, tablets and computers equipped with built in or external modems, therefore the following program is implemented in JAVA.

The project implemented an estimation method for link capacity based on UDP burst measurements into the MONROE system. The validation process was done in a special anechoic chamber with a fully operational live network base-station. This allowed for a fair comparison between the new and existing methodologies, e.g., iPerf3 baseline measurements, under different load conditions. The validation in the real-world was achieved with two drive tests campaigns in Austria. The first drive test focused on the city of Vienna to analyse high to very-high datarates. The second drive test featured all of Austria along the high level road network, e.g., high ways. In both test the module proved to be stable and reliable in terms of operation as well as in terms of results compared to state-of-the-art measurement methodologies. The MONROE project is now equipped with a new module for volume lightweight and fast bandwidth estimation suitable for crowdsourced measurement systems. The validation and the drive tests show that speeds up to an unloaded LTE cell can be measured with as few as 2MB per test. Even low volume tests of as low as 60kB are suitable to verify link capacities of 10Mbit/s.

Whereas not an issue for high data volumes and long test durations, in the case of low volumes and short durations more careful processing – considering individual bursts of data – is crucial. In addition to this the measurement in the reference cell with a tariff limited SIM card showed that, in the network under test, the measurement result was not impacted by shaping due to the user tariff linked to the SIM card.