Research of a BLE and/or ANT architecture - student project

[quentinDstl]

Hello,
As part of a student project I am looking to learn about Nordic technologies for BLE and ANT.
The goal of our end of study project is to provide an extremely accurate tracking of a soccer game.

One of the solutions is to locate all the players by equipping them with BLE transmitter and triangulate them to have their relative position.
In view of the lack of precision of the BLE localization protocols, we planned to make them send timestamps to each other and process the information on an external device.
We want to use the AoA and AoD present in the 5.1 to complete the data with the rotation angle of the players.

We are wondering if all this is possible and if you have any advice on how to realize the project ?

Is it possible to use the direction-finding protocol (to have the principles of AoA and AoD) in parallel with the BLE (to reduce the energy consumption) and the mesh (to be able to compensate for the high distances that the information will have to travel by relaying it between the other devices).

We intend to recreate our own stack to be able to use these different advantages of bluetooth at the same time but are for the moment in the blur for this possibility.

Thanking you for all the answers and wishing you a pleasant day

[tesc]

Hi,

Positioning and tracking is an interesting, albeit difficult field of study. My main recommendation would be to study state of the art within the field, and quantify what you want to achieve within this space.

Here are some pointers regarding what exist today:

Bluetooth 5.1 AoA and AoD relies on antenna arrays of typically 8 or more antennas and a total size of the array at around 20 by 20 cm. With a wavelength of around 12.5 cm for 2.4 GHz signals (such as used for BLE and ANT) an antenna array size of this size is typically what you would need for getting precision down towards 5 degrees accuracy. The BLE AoA and AoD technologies relies on one end of the link to have such an array, while the other end use a single antenna. This means if used for tracking, the array can be stationary while the tracker can be a much smaller device on the object being tracked. Please note that with this setup, the rotation of the tracked object is not known. I'd recommend using RF for the location, and rather use a compass sensor for getting the rotational orientation of the object being tracked.

AoA and AoD are, in fact, used in conjunction with BLE traffic, since what it does is to add a constant carrier at the end of normal BLE packets. This constant carrier is received (or sent) by multiple antennas, so that the signal  received by the radio at the receiver end is "broken up" into multiple segments with a phase shift between each segment. You get a phase shift because the traveling distance of the signal is different to the different antennas, and so they are "out of phase" relative to each other. Analysis of these phase shifts (with knowledge of the spatial configuration of the antenna array and which antenna was activated when) gives you the direction from which the signal came.

One technology that you haven't mentioned, is distance measurement. We do have a distance measurement toolkit in nRF Connect SDK. See our webinar Measuring distance with the Nordic Distance Toolbox for details on this technology. In particular, round-trip-time might prove useful for distances longer than around 15-20 meters, although our current accuracy is +/- 5 meters. For closer than 15 meters distances we see accuracy around half a meter.

The size and environment of a soccer field might be an issue. Even if you have line-of-sight, which is good, a stadium is often full of mobile devices and fixed equipment utilizing the 2.4 GHz band. There may also be issues with multi-path.

Regards,
Terje