Exploring Non-Idealities in Real Device Implementations of Bluetooth Mesh

Background

This work is based on multiple technical publications of which some shall be highlighted in this section.

The underlying Bluetooth (Low Energy) specification can be found in (Bluetooth Special Interest Group, 2016). De Cerio et al. (2017) describe non-idealities that occur when Bluetooth Low Energy is used on real chipsets. Their findings of scanning gaps in the Bluetooth Low Energy scan process is an important foundation for the presented work. They experimentally measure those scanning gaps on multiple platforms including the Nordic Semiconductor nRF51822 and formulate a mathematical model of the impact of those non-idealities on the Bluetooth discovery process performance.

Bluetooth Mesh is specified by the Bluetooth Special Interest Group in three documents: the profile (Bluetooth Special Interest Group, 2018c) which defines the stack, models (Bluetooth Special Interest Group, 2018b) that are pre-built interfaces to common applications and device properties (Bluetooth Special Interest Group, 2018a) that are required for the Mesh model specification. Furthermore, in (Hollander, 2018) one of the key applications is mentioned: smart lighting. Not only is lighting itself a good example for an application of Bluetooth Mesh but lighting devices may also serve as a natural infrastructure grid in buildings to allow multi-hop communication with otherwise sparingly seeded nodes.

However, as of now, there is little scientific research on Bluetooth Mesh.