Low-power wireless network with real-time guarantees for edge-cloud applications

Abstract

The goal of this project is to explore the feasibility of building a scalable & easy-to-deploy real-time LoRa testbed, made from multiple units of Raspberry Pi (RPI), where each RPI manages its own set of LoRa radios. This project is motivated by the lack of concrete large-scale LoRa testbeds that effectively integrate LoRa communications into the real-time world. The paper introduces how the idea of using RPI came about and why it should work in theory. The paper then carries out experiments on a component of the large-scale testbed, to evaluate the feasibility of the said component based on performance metrics such as RSSI, SNR, PLR and the ability to carry out millisecond-accurate transmissions. The performance metrics are also used to explore the impact of using different combinations of spread factors and transmission frequencies, as well as making comparisons between time-division multiple access (TDMA) and carrier-sense multiple access (CSMA) approaches. The results show that with the right parameters configured, the system can achieve stable and low-latency communications, proving some feasibility to operate under real-time situations. Future work includes giving each RPI control over more radios, carrying out true parallel transmissions, and finally integrating multiple RPIs for a more complete large-scale real-time LoRa testbed.