InnerWeb
Skip to main content

FAQ: LoRa at an industrial plant. Questions from the IT department

The IT department rarely objects to LoRa radio as such. It wants to know what the radio does to the network, the spectrum and the data. We collected the questions that come up before the first gateway is approved at a plant, and we answer them with figures from the watch and gateway configuration.

about 0.2%
airtime of one watch at a 120 s interval
Source: karta
10%
airtime limit in the 869.4-869.65 MHz sub-band
Source: karta
AES-128
frame encryption, 8-byte authentication tag
Source: karta

Spectrum and regulation: does the radio need a licence

The watch and the gateway operate in the 869.525 MHz channel, within the 869.4-869.65 MHz sub-band, where the power limit is 500 mW ERP and the airtime limit is 10%. This is a general-use band in Europe, so the plant user does not apply for an individual frequency licence. The user is responsible for keeping the devices within the parameters of the band. The IT department should receive these parameters in writing before the purchase.

The watch transmits at +14 dBm with an SF9 spreading factor and a 125 kHz channel width. +14 dBm is about 25 mW, many times below the permitted 500 mW ERP, which leaves a margin for a gateway antenna with higher gain.

Will the radio interfere with Wi-Fi or OT measurements

LoRa does not operate in the 2.4 GHz and 5 GHz bands, so it does not share channels with Wi-Fi or BLE. The watch scans BLE for about 3 s per cycle and does not keep connections open that could load the infrastructure.

If other devices already operate in the 868 MHz band on site, for example meters, sensors or a mesh network, they must be included in the frequency plan. We cover this topic in a separate guide on channel planning.

Airtime and channel occupancy

The most common question is whether the watches could use up the airtime limit. One frame lasts about 230 ms on air. With the default 120 s interval this means about 0.2% airtime for a single device, far below the 10% limit.

The limit applies to each transmitter separately, but the IT department rightly asks about the occupancy of the channel itself. Example calculation: 100 watches sending a frame every 120 s occupy the channel for 23 s in every 120-second window, which is about 19%. This value is planned for deliberately and is not ignored. Shortening the interval to 60 s raises it to about 38%, so frame collisions become frequent.

  • The legal limit of 10% airtime applies to one transmitter, and in our configuration it is respected with a margin.
  • Channel occupancy by the whole watch fleet grows linearly with the number of watches and falls with a longer beacon interval.
  • The beacon interval is set centrally in the panel within the range of 30-3600 s and is sent back to the watch in the system response.

Channel occupancy by 100 watches per beacon interval

Channel occupancy by 100 watches per beacon intervalBar chart: channel occupancy grows from 3.8% at a 600 s interval to 76.7% at a 30 s interval020406080%30 s: 76.7 %76.7 %30 s60 s: 38.3 %38.3 %60 s120 s: 19.2 %19.2 %120 s300 s: 7.7 %7.7 %300 s600 s: 3.8 %3.8 %600 s
Source: Example calculation: 100 watches, 230 ms frame, occupancy = 100 x 0.23 s / interval

Security: encryption, token and time

Frames of the InnerWeb v2 protocol are encrypted with the AES-128-CCM algorithm, with an 8-byte authentication tag and counters that prevent replay of a captured frame. For the IT department this means that listening to the radio does not reveal the frame content, and an old frame that is replayed is rejected.

The gateway validates the TLS chain of the connection to the system and can send a gateway token. Once the token list is set, the system enforces the token, so a substituted foreign gateway will not be accepted. The gateway is the time source for the watch: it synchronises via SNTP and includes the UTC time in each acknowledgement, so the IT department must give it access to a time server.

What the IT department receives for the security documentation

The standard set covers the description of the algorithm and key length, the replay protection mechanism, the way the gateway is authenticated and the outgoing connections required. All of this supports the risk assessment and does not replace the plant's own assessment.

Plant network: what the gateway needs from IT

The LoRa gateway translates watch frames into HTTP requests to the system. It confirms receipt of a frame within about 250 ms, before the request is executed, and adds the RSSI and SNR of the link to the acknowledgement. For IT this is outgoing traffic from one device, with no incoming traffic from the internet.

Not every gateway on site has Wi-Fi. Data from all gateways must reach the one that has internet access. The default variant is a repeater gateway with a hop counter of 2-3 and frame deduplication. Where a gateway is far away but has power and GSM coverage, an LTE module is selected, and in halls with Ethernet cabling or PoE, a wired connection is used.

Working outside beacon range and during SOS

When the watch hears no beacon, the frame is still sent. The system substitutes the last known location, so battery status, state and inactivity reach the system even from places without beacon coverage. IT therefore does not need to treat the lack of a beacon as a lack of supervision over the device.

During an SOS, the gateway sends a second frame with the result after completing the HTTP request. The watch then shows that the alarm reached the system, which matters for the user's trust in the button.

Video

Frequently asked questions

Does LoRa require a frequency licence?

The 869.4-869.65 MHz sub-band is general-use in Europe, and the user is responsible for staying within 500 mW ERP and the 10% airtime limit. The watch transmits at +14 dBm, which leaves a large margin.

Can someone eavesdrop on or replay the frames?

Frames are encrypted with AES-128-CCM with an 8-byte authentication tag, and counters cause a replayed frame to be rejected. Radio eavesdropping does not reveal the content.

How do we protect the system against a foreign gateway?

Once the token list is set, the system enforces the gateway token, and the gateway validates the TLS chain of the connection. A gateway without a valid token will not be accepted.

What about a gateway that has no Wi-Fi?

By default a repeater gateway passes frames to the gateway with internet access. For remote points with power and GSM coverage, an LTE module is selected.

IT questions before a LoRa pilot

We will prepare a one-page description for your IT team covering the band, encryption and the required gateway connections. Let us arrange a short technical call.

Let's talk