Multipath XR2 Error Correction, Indoor Asset Tracking

LoRa in Multipath: Theory vs. Experiment

Recently, we came across this great paper (link, free PDF) about modeling the impact of multipath on LoRa modulation.  Specifically, it models Rayleigh flat fading, which is observed in urban and indoor use-cases.  We had just finished testing LoRaWAN vs. Haystack’s LoRa XR2 in an environment with heavy multipath interference (an airport parking garage).  The results were so one-sided that we wanted to find any research that might either validate our results or, potentially, shed light on something we were doing wrong in the experiment.  Fortunately, we found just the paper, and it corroborates perfectly with the experimental results we achieved.

The presentation below describes our findings.  If you’re too impatient to read it, here are the basic findings:

  • Haystack XR2 yielded an observable 30dB sensitivity improvement over LoRaWAN in the Rayleigh flat fading environment, and this matches the theoretical model.
  • In such environments, system integrators can use higher data rates with Haystack XR2 to trade-off QoS vs. energy efficiency vs. channel density in such urban or indoor environments.