LPWAN Day is an annual conference for university research in the field of LPWAN. This year’s edition was held in Pau. The event provides an opportunity to review the state of research on technologies such as LoRaWAN, Mioty, Wirepas, and satellite experiments, fostering technical, detailed, and inspiring exchanges.

This year, about sixty participants attended, mainly from academia but also from industries like Semtech, Wirepas, Schneider, and Kineis. As in every edition, we had many very in-depth and passionate discussions. The organization was flawless, and the hospitality in Pau was exceptional. However, I must admit that our visits to Glacier Giorgios may have skewed our perceptions.

In summary, it was two wonderful days, with beautiful weather and exceptional people. Here are my key takeaways.

Mioty

Last year in Grenoble, we already discussed Mioty, a technology I haven’t mentioned much because I haven’t had the chance to experiment with it. However, I was promised a gateway in Munich, so I remain hopeful. Last year, the Fraunhofer Institute gave a presentation, showcasing the technology. I had previously criticized its aggressive marketing towards other LPWANs, lacking technical accuracy.

This year, we had three presentations on Mioty. Two were by proponents of the technology, and the third by Martin Heusse offered a scientifically objective, well-documented, and enlightening perspective. It turns out Mioty is advancing with a standardization of its downlink classes similar to LoRaWAN’s A, B, C, and a unique Z class (no downlink), highlighting a crucial aspect: the lack of downlink capability. Mioty, like Sigfox and LR-FHSS, is an asymmetric communication technology requiring complex, costly, and energy-intensive SDR (Software Defined Radio) for reception.

For now, Mioty can be used in uplink-only mode with a standard MCU like STM32WL and an SDR concentrator, costing about €200. The second presentation highlighted Mioty’s use in wildlife protection, a recurring theme in IoT, previously illustrated by Sigfox and LoRa. While always attractive, it no longer distinctly demonstrates the advantages of one solution over another.

Martin Heusse’s presentation, was very enlightening about Mioty’s claims, particularly regarding its capacity. To summarize Martin’s presentation: Mioty’s sensitivity should match LoRa’s SF12, but using SDR incurs a significant sensitivity cost (40%). While Mioty claims higher throughput (2.38Kbps), a fair comparison should consider data duplication (3x) and synchronization bits, reducing real throughput to 529bps, comparable to LoRa SF12 at 250KHz.

In terms of network capacity, Mioty claims theoretically supports 26,000 times more devices with a 99% success rate. However, practical application reduces this to about one factor of magnitude more, considering simple message repetition in LoRa and spectral occupancy. For optimal message reception in LoRa, use repetition rather than acknowledgment, as practiced by Sigfox and Mioty.

In conclusion, Mioty offers a significant capacity increase with similar throughput to other solutions, considering its spectral occupation, almost reaching 1.4MHz in its most capacitive use. It has advantages in frequency and time diversity compared to LR-FHSS but requires 1/3 more bits for synchronization. There is no magic in science, everything is a compromise.

LR-FHSS

We saw the first results of open-field (with helium balloons) and urban experiments, comparing LR-FHSS with LoRaWAN transmissions. Using Kerlink stations, experiments showed similar functionality, with ranges exceeding 80km in open fields. The technology uses 200KHz channels (OCW) for packet transmission, offering frequency diversity over eight time slots, received on a standard LNS like ChirpStack. Like Mioty, LR-FHSS doesn’t support downlink, but chips like the LR1120 can receive LoRa SF12 downlink.

Experiments showed an average reception rate of 75% (without repetition) across various data rates and transmission powers (6-16dBm). Urban tests indicated a slight advantage for LR-FHSS over LoRaWAN. While terrestrial experiments have begun, LR-FHSS is mainly used for satellite applications due to its better performance for horizon coverage.

A consensus emerged: the name of this technology is unpronounceable!

Relay

LoRaWAN relays were recently standardized, but the current standard isn’t fully functional. Significant work is underway within the LoRa Alliance. Estimations suggest a LoRaWAN relay’s battery life (17Ah LS33600) is about 10 years for one channel and 7.5 years for two channels, which seems reasonable.

New Space IoT

Several topics on satellite IoT were discussed, with significant off-topic discussions, particularly with Kineis. Fabien Ferrero’s team presented an autonomous device, using solar energy, communicating with Echo Star (geostationary LoRaWAN LR-FHSS), and featuring a camera and AI module for detecting forest fires. A super project, functional and inspiring.

Similarly, the Pau community’s presentation on watercourse monitoring was very interesting, potentially inspiring many municipalities.

ChirpStack & TTS

Oana Iova conducted a study comparing ChirpStack and The Things Stack (TTS) in downlink management. The study found that downlink duty cycle management isn’t considered in gateways or LNS. TTS favors RX2 (beneficial in Europe for better link budget at the cost of SF12 emissions) while ChirpStack prioritizes RX1. ChirpStack pushes all downlink requests to the gateway, potentially overloading it without user feedback, whereas TTS schedules downlinks, rejecting them at the LNS level but improving delivery and providing user feedback. There has been much more from the talk, but better reading the paper to go on details.

Other Topics

Presentations included SCHC, bridging IoT data and IPv6 transparently, continuing with open-source (openschc) and industrial solutions like Actility’s IPCore.

I also discussed Helium and existing datasets for research on massive radio exchanges and coverage proofs.

Another presentation covered modular firmware approaches for more efficient FUOTA updates.

Conclusion

The conference was rich in content, not to mention the off-topic discussions. Here, I’ve highlighted the main points that impacted me. For more details, visit the conference website: LPWAN24. Kudos to Congduc Pham and the entire team for the organization.