Resource Constraint Networking (RCN)

Problem Space

The general assumption of today’s Internet service providers and users of such services is that the Internet is always capable of delivering the services desired by the users. For instance, accessing a video webpage, such as, youtube or dailymotion, requires an existing, stable and bandwidth-wise suitable Internet access. A failure to fulfill the service requirements would render the access to the webpage useless, but this wouldn’t cause a real harm.
However, there are scenarios where a stable access to the Internet can be crucial: industrial or medical applications rely on a stable Internet access in order to function properly. One industrial application is self-driving cars, where the car’s computer will need real-time information about the route’s condition and computing support to fulfill the goal of self-driving.

On the other hand, there is the reality of lack of network coverage of cellular networks in Germany, but potentially also in other areas. This lack of coverage has multiple reasons, but in perspective this will leave many geographic areas with limit or no cellular network coverage, despite the fact that some industrial applications require a constant, stable network connection.

Potential Remedies

There are a number of different remedies to solve this challenge, but one approach relies on the end hosts in such scenario. It is possible to equip an end host, i.e., a car, with multiple cellular network adapters and let these adapters connect to different mobile network operators. The prospect is that not all operators will offer the same level of limited network coverage, but that a combination of multiple operators can provide a gain in network connectivity. The end host has either to route its traffic over the adapter that has best connectivity or to aggregate the connectivity of multiple adapters.

Scope of the Project

The project has the goal to determine whether existing solutions are suitable to deal with resource constraint networks or if new solutions are required. If existing solutions are suitable, the project will explore how to use them in real usage scenarios.

The main assumption is that there will be always spots or larger areas with lack of suitable Internet connectivity or no Internet connectivity at all.