Powerpoint exploring the locations used in television show Time Clash
Impact of temperature on LoRa networks
1. Carlo Alberto Boano
Graz University of Technology, Austria
cboano@tugraz.at
Impact of Temperature Variations
on the Reliability of LoRa
– An Experimental Evaluation –
LPWAN London Meetup #19 @ Digital Catapult - Wednesday, July 3, 2019
+ = ?
2. Impact of Temperature Variations on the Reliability of LoRa
Reports from Outdoor WSN Deployments2
Meteorological weather station
Uppsala, Sweden
(Wennerström et al., 2013)
Deployment in Botanical Gardens
Phoenix, AZ
(Bannister et al., 2008)
Temperature affects wireless communications
• Received signal strength of packets decreases at high
temperatures, with daily fluctuations up to several dB
• Worst connectivity during summer and during daytime
3. Impact of Temperature Variations on the Reliability of LoRa
Reports from Outdoor WSN Deployments3
Temperature affects wireless communications
• Received signal strength of packets decreases at high
temperatures, with daily fluctuations up to several dB
• Worst connectivity during summer and during daytime
• Reason: temperature affects low-noise & power amplifiers
(experiments by Bannister et al., 2008 on TI CC2420 radio)
High temperatures increase the resistance of conductors,
while reducing the pass-through current → reduced SNR
• Even worse: airtight, waterproof, and IR-transparent
enclosures, cause diurnal/seasonal differences
up to 50 and 85ºC, respectively!
4. Impact of Temperature Variations on the Reliability of LoRa
Reports from Outdoor WSN Deployments4
Result: low performance at high temperatures!
This evidence is mostly based on IEEE 802.15.4 radios!
5. Impact of Temperature Variations on the Reliability of LoRa
Does this also
apply to LoRa?
5
6. Impact of Temperature Variations on the Reliability of LoRa
Experimental Setup6
We use TempLab, a testbed infrastructure capable
of reproducing real-world temperature profiles
7. Impact of Temperature Variations on the Reliability of LoRa
Experimental Setup7
TempLab allows reproducible distributed experiments
• Temperature can be adjusted in the range [-5, +85]ºC
• High accuracy (up to 0.1ºC when reproducing a trace)
• Isolate the effects of temperature changes on specific
nodes (e.g., on transmitter / receiver only)
C.A. Boano et al. TempLab: A Testbed Infrastructure to Study the Impact of Temperature on Wireless Sensor Networks. In Proc. of IPSN‘14.
8. Impact of Temperature Variations on the Reliability of LoRa
Experimental Setup8
Tested two LoRa platforms
• LowPowerLab Moteino MEGA
(HopeRF RFM95 radio)
• ST Nucleo L073RZ
(Semtech SX1272 radio)
Star network with a sink (node 0) periodically
broadcasting messages with a 5-bytes payload
• Nodes 1 and 2 at the edge of sink’s communication range
• Nodes 3 and 4 closer to the sink (higher signal strength)
9. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results9
1. Link disruption: an increase in temperature can
compromise a LoRa link completely
PRR(%)
10. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results10
2. Impact is highest when both transmitter and receiver
are heated at the same time (TX+RX)
PRR(%)
11. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results11
3. Attenuation of received signal strength as a function of
temperature can be modeled (e.g., with linear function)
(≈ 3-4 dB in the temperature range [0-50]ºC)
12. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results12
4. RSS attenuation is platform-specific
Semtech SX1272 radio HopeRF RFM95 radio
13. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results13
5. Changing PHY settings helps mitigating the impact
A lower bandwidth increases the robustness of the link
14. Impact of Temperature Variations on the Reliability of LoRa
Experimental Results14
A higher spreading factor increases the robustness of the link
5. Changing PHY settings helps mitigating the impact
16. Impact of Temperature Variations on the Reliability of LoRa
Conclusions16
1. Is the reliability of LoRa affected by temperature?
• Yes: an increase in temperature can destroy a LoRa link
• High temperatures increase packet corruption and loss
2. Is the impact platform-specific?
• Yes: it varies based on the employed radio
3. Can the impact be mitigated?
• Yes, with a proper selection of PHY settings
• It can be mitigated by each node autonomously
17. Impact of Temperature Variations on the Reliability of LoRa
Implications17
Careful deployment of nodes is necessary
• LPWANs are mostly deployed outdoors
• Deploy always during the warmest time of the day / year
• Orchestrator node must be shielded from sunlight!
(LoRa networks = star networks)
Let each node monitor its on-board temperature…
• Understand if the low performance is temperature-related
… and adapt PHY settings accordingly!
• Autonomously switch to a lower bandwidth, a higher
spreading factor, and/or a higher coding rate
18. Carlo Alberto Boano
Graz University of Technology, Austria
cboano@tugraz.at
Impact of Temperature Variations
on the Reliability of LoRa
– An Experimental Evaluation –
Work presented at the 7th International
Conference on Sensor Networks last year
LPWAN London Meetup #19 @ Digital Catapult - Wednesday, July 3, 2019