Eertmans, Jérome
[UCL]
Oestges, Claude
[UCL]
Louveaux, Jérôme
[UCL]
Similarly to other technologies, Bluetooth has received increasing interest over the past 20 years. In par- ticular, the Bluetooth Low Energy (BLE) market undergoes constant growth for more than a decade, with billions of device units shipped every year. Initially thought to be simple and low power, BLE is getting more complex year after year as new applications emerge. The state-of-the-art products require multi-connectivity and high-quality link while staying low power. This leverage some severe design challenges that engineers must solve. After highlighting the complexity related to modern challenges, this document proposes a novel algorithm for multiple connections tasks scheduling. Another very challenging mission is coexistence interfer- ence. BLE uses the same frequency bands as Wi-Fi, a prevalent technology in most homes, companies and public facilities. As such, channel assessment, a mechanism to evaluate the quality of the different frequency bands, is essential to avoid interference. Current implementation sees Wi-Fi as the source of interference, but not the other way around. An enhanced channel assessment is proposed to detect the strongest Wi-Fi signal disregarding its signal strength and the number of active Wi-Fi channels. After detection, the BLE channels overlapping with Wi-Fi’s are removed from the channel map to avoid interference. Finally, the two contributions discussed reveal encouraging results for real-case applications. Task scheduling always finds an excellent solution in less than 100 ms. Under several fairly realistic assumptions, this time can be reduced to under 10 ms, or even 1 ms in the best cases. As for Wi-Fi signal detection, this improvement adds less than 20 µs to the current execution of the Channel Assessment. This method, which is robust to outliers, makes it possible to detect, in a few milliseconds, the location of the strongest Wi-Fi channel in a crowded spectrum with a minimum accuracy of between 2 MHz and 10 MHz depending on the case.
Bibliographic reference |
Eertmans, Jérome. Radio activity optimisation for Wireless Personal Communications. Ecole polytechnique de Louvain, Université catholique de Louvain, 2021. Prom. : Oestges, Claude ; Louveaux, Jérôme. |
Permanent URL |
http://hdl.handle.net/2078.1/thesis:30592 |