Public Transport

The business value of great QoS

Being able to access the internet can be a decisive reason to choose public transport over travelling by car. If mobile network coverage was good, people could spend their time on board trains or buses communicating or using multimedia services for work or entertainment.

From a railway operator's perspective, these demands are a key commercial factor. If ticket sales could be increased by providing a good mobile experience, investments in this field would easily be justified.

However, investments in infrastructure and services in the realm of QoS/QoE should be based on metrics that allow to exactly evaluate the business value of a service and thus provide orientation how to best allocate investments.

Standard KPI insufficient in public transport

Current standard KPI, however, have several shortcomings when it comes to predicting the usability of the network in a public transport setting.

A single number KPI like a drop rate reflects only a single aspect of the user experience, neglecting any information about a particular route or service usage. Its informative value is therefore limited. The drop rate, for example, doesn't allow predictions regarding the expectable length of a call that starts at any given location. It therefore fails to answer typical user questions like

Will my call hold and will speech quality be sufficient over the duration of my call?
Can I access web sites or use social media comfortably during my journey and can I use e-mail or other means of communication?

Specific KPI for Public Transport

In public transport situations, use cases are strongly tied to motion. This movement is one-directional and – especially on board trains – characterized by a more or less constant speed. Also, a train or bus always runs on the same route and on a given timetable. The use case is therefore repeatable in both space and time. Consequently, test case and use case are very similar, and the macroscopic network topology becomes a property of the use case.

A more holistic approach offers several advantages and mirrors the spatiotemporal characteristics of public transport:

It allows assertions about the probable duration of a call started at any location.
It is call length independent: The results from shorter calls can be used to make predictions about the results of longer calls.
It reduces the data volume required to arrive at meaningful results, because it provides more data points, not only endpoints.
It provides a much more detailed view.

Orientation for smart investment decisions

The KPI model can be filled with different types of measurement data. Usable metrics are possible with only a few passages on the route, while additional measurement data fine-tunes the assertions.

The model can be used to collect information about the network with a single measurement. On top of that, it can be used to set up constant monitoring based on, for example, autonomous measurement systems, to continuously update the knowledge about the route. It is also possible to integrate data from other sources, e.g onboard WiFi routers or repeaters, or even crowdsourced data.