21 Mars – Thesis defense - Mohamed Amine Ben Temim

10 h Amphi Jean-Paul Dom - research unit IMS | Building A31 (Talence campus)

Low-Earth-Orbit satellite communications using LoRa-like signals.

Connecting a device to the Internet is nowadays possible through several communication technologies. However, a limited area of the planet is currently connectable to the Internet via terrestrial cellular networks.
Therefore, with the development of satellite Internet of Things (IoT) in recent years, it is possible to provide reliable communication services for places where there are no terrestrial networks. Thus, satellite IoT is a very ambitious sector allowing to blanketing the Earth with reliable and ubiquitous coverage.
In terms of power, propagation delay, and coverage, low-Earth orbit (LEO) satellites are more suitable for IoT communications than other types of satellites.
LEO satellite communications are currently facing two major challenges. The first one stems from their high speed yielding significant Doppler effects. The second challenge is the huge number of devices that could be connected to the latter satellites due to their field of view (FoV), which leads to a high probability of packet collisions.  In this thesis, we deal with IoT communications with LEO satellites using low power wide area (LPWA) technologies in unlicensed bands. Typically, we focused on LoRa-like communications using chirp-based waveforms.
On one hand, our first contribution is to propose several synchronization algorithms allowing to accurately decode LoRa-like signals received with random arrival times and with significant Doppler effects especially the Doppler time-variation.
On the other hand, for energy efficiency reasons, most LPWA technologies in unlicensed bands adopt uncoordinated channel access schemes which leads to an increase in the probability of packet collisions, given the huge number of objects that can be connected to a LEO satellite.
Thereby, our second major contribution consists in proposing novel approaches to decode interfering LoRa-like signals in uplink and downlink contexts, based on the successive interference cancellation (SIC) algorithm.

Event localization