Whitepapers

LoRaWAN ™ and LoRa Alliance ™ are marks used under license from the LoRa Alliance ™ www.lora-alliance.org GEOLOCATION WHITEPAPER On average, the timestamps are late. Note that the timestamp errors may be negative, but they are never smaller than -1/bandwidth (never lower than the basic resolution of the system). We can classify the timestamp errors with a bias, which is the average error, and a spread around this average. Bias increases and spread widens as the propagation environment degrades. There are various ways to mitigate timestamp errors. • Repetition of frames at different frequencies • Antenna diversity at the gateway (typically two antennas) • Higher-density gateway deployment, which increases the number of available samples (frame receptions) and increases the chance of line-of-sight measurements, thus increasing the accuracy of the TDOA • Lower latency frame timestamping at the gateway • Incorporation of out-of-band propagation error corrections to mitigate multipath (simulations, predictions, calibration or fingerprinting) The geolocation solver is designed to mitigate multipath. The solver selects a candidate set of timestamps to be processed, ignoring the rest. The solver then aggregates the remaining data and solves for the end-device location. Multipath propagation fundamentally limits the accuracy of the system, but gateway-deployment geometry also plays a significant role. 4.3 Impact of Deployment Geometry As with other radio navigation systems (e.g., GPS, LORAN), the accuracy of a LoRaWAN geolocation position fix depends on the placement of the gateways vs. the end-device. The metric used to determine the quality of gateway placement is the Geometric Dilution Of Precision (GDOP), which is a measure of the "goodness" of the receiving gateway's relative geometry. Figure 4-2: Statistics of Timestamp Errors

• Cover