top of page

LoRaWAN Downlinks: Prove a Queued Setting Reached the Sensor

24 hours ago
4 min read

A remote sensor’s reporting interval has been changed in a console, but its messages still arrive at the old cadence. Before sending the command again, establish what actually happened: was the request accepted, scheduled for radio transmission, received by the device, and applied by its firmware? Keep these as separate checkpoints in the commissioning record.


Class A makes the next uplink important


The Things Network’s device-class guidance explains that a Class A end device opens short receive windows after an uplink. A downlink cannot simply reach a sleeping Class A sensor whenever an operator presses Send. If the available receive opportunity is missed, another uplink is needed for a subsequent Class A opportunity.


That relationship matters when changing a long reporting interval remotely. Plan the observation period around the current behaviour, rather than assuming the requested new interval already applies. Exact receive delays and supported commands depend on the deployed configuration; this guide does not prescribe universal timing or command bytes.


Dragino wireless LoRaWAN temperature and humidity sensor shown in a catalogue product photograph
Dragino wireless LoRaWAN temperature and humidity sensor from the ProSense catalogue. Product photograph, not a customer installation.

ProSense’s Dragino wireless LoRaWAN temperature and humidity sensor is listed with Class A communication. Use its exact hardware and firmware documentation when choosing a configuration command. The photograph illustrates the type of monitoring device involved; it does not establish that a particular command is supported by every Dragino sensor or that it integrates directly with ProSight.


Separate scheduling evidence from application evidence


The Things Stack troubleshooting documentation distinguishes events produced by its Application Server, Network Server and Gateway Server. For example, as.down.data.receive and as.down.data.forward relate to application-side processing; ns.down.data.schedule.success records successful network scheduling. Gateway-side gs.down.tx.success indicates that the gateway scheduled the downlink for transmission. These events help locate a problem, but they are not a read-back of a sensor’s configuration.


Record the event names and timestamps from the actual platform version. When scheduling fails, preserve the failure detail rather than treating the absence of a changed dashboard value as the only symptom. The Things Stack also identifies slow server-to-gateway connections as a possible reason for a downlink arriving too late.


A practical commissioning worksheet


Start with the device identity, current firmware, network-server deployment and intended configuration change. Record the documented application port, encoding and payload selected for that exact device. Have the responsible integrator verify these before transmission. A syntactically valid request can still target the wrong device or express the wrong setting.


For each authorised change, keep one record linking the request time, command identifier, queue state, relevant uplink, scheduling events and device-side evidence. Where the firmware provides a documented settings query or configuration response, retain it. Where it does not, agree how behaviour will be assessed and label that evidence accordingly.


Set a review time and an escalation owner before the test. The review time is an operational checkpoint, not a promised delivery deadline. Avoid adding repeated requests merely because the dashboard has not changed yet; first inspect the outstanding request and its evidence. Check the platform’s documented queue behaviour before changing or replacing pending commands.


Worked example: a request made just after an uplink


Assume a hypothetical Class A sensor sends regular readings every 30 minutes, at 10:00 and 10:30. At 10:02 an authorised operator queues a supported request to change the interval to ten minutes. Assume the 10:00 receive windows have already closed, and no other uplink occurs before 10:30.


Under those assumptions, expecting a changed cadence at 10:12 is premature. The request must first have a suitable receive opportunity after a later uplink. At 10:30, inspect the actual scheduling and gateway events. If those show success, look for the documented device response or subsequent behaviour that supports the change having taken effect.


A later message at 10:40 would be consistent with the requested interval, but one timestamp alone does not establish the complete configuration. An alarm, manual trigger or delayed data delivery could complicate the interpretation. Keep the original sensor timestamps, server reception times and message types where available, and compare several observations. These times are illustrative, not a performance claim for the photographed product.


Close the change with evidence


Common mistakes are timing the test from the requested new interval, treating a queued request as completed, and copying command bytes from a different model. Another is combining device timestamps and server arrival times without labelling them. A handover should state the intended setting, strongest available evidence, remaining uncertainty and who accepted the result.


If the evidence stops at network scheduling, report that exact limit. Send ProSense the sensor model, firmware version, current and requested intervals, relevant event timestamps and redacted payload details. Exclude security keys. Application-specific integration must be checked separately; remote monitoring configuration should not be relied upon as a time-critical safety function.


Sources and product information





Need help with your application?


ProSense Instruments - explore instruments, sensors, data loggers and test equipment.


ProSense Solutions - discuss complete monitoring and automation systems.


ProSight Software - explore cloud dashboards, alarms and reporting. Device integration must be confirmed for your application.



Project enquiries: sales@prosense.com.au. Software enquiries: info@prosightsoftware.com.

Comments


bottom of page