I’m experiencing intermittent vehicle disconnections during AUTO missions and I’m trying to understand whether this is related to my antenna setup or radio configuration.
Setup
Two communication options (both Blue Robotics): omnidirectional antenna and panel antenna
Base station on a tripod at ~1–1.5 m height
Distance from vehicle: ~30–50 m
Same location and conditions for both tests
Omnidirectional antenna
The connection continuously drops and reconnects during the mission.
In Cockpit I see:
Vehicle disconnected / connected
Auto mission changed, restarted command
In MikroTik WebFig the signal strength is around –69 dBm.
Panel antenna
With the panel antenna in the exact same setup:
Connection is stable
No disconnections
Mission runs normally
Question
Since the omni is rated for a longer range (~250 m), I would expect it to work reliably at this short distance.
Is this behavior expected close to the water surface?
Is –69 dBm enough for a stable MAVLink link in this scenario?
Besides signal strength, which parameters should I monitor to detect link instability?
What practical changes or field tests would you recommend to improve stability with the omnidirectional antenna?
I’ve attached screenshots of the Cockpit log and the MikroTik registration tab.
Thanks in advance — I’m happy to run additional tests if needed.
Hi @MatteoBucci -
The Omnidirectional antenna range may be able to achieve 250m in some environments, but it is certainly not always going to! Antenna elevation is critical - is the antenna on the BlueBoat on deck or elevated with a mast?
What is the signal strength with the panel antenna?
Are there lots of other WiFi networks around the area you’re operating?
That said, connection is not typically lost until around -80dB, so what you’re experiencing isn’t normal. Have you ever operated either radio, BlueBoat or BaseStation, with no antenna connected? This can damage the radio - swapping antennas should occur with the system powered off…
I’ve not come across the error " * Auto mission changed, restarted command" before. Is your failsafe parameters setup to continue on auto ?
1. With the panel antenna, at approximately the same distance, I was getting around -54 dBm as shown in the figure.
2.The GCS is always mounted on a tripod at about 1.5 m above the ground.
3. In the third picture you can see that the area near the pond is fenced. During navigation the GCS is positioned there (on the tripod at 1.5 m, not on the red box). Could the presence of the metal fence affect the link?
That said, I experienced the same disconnection issue in other test locations as well.
4. I don’t think there are other Wi-Fi networks nearby. The only other system in the area is the GNSS, which communicates with its base via radio.
5. I always powered everything off before swapping antennas.
6. The failsafes are set to continue the mission, as shown in the figure.
7. When the disconnection happens and the link comes back, in addition to the message
“Auto mission changed, restarted command” I sometimes also get “mission download failed”, as shown in the last figure. Do you have any idea what might be causing these messages?
The mission continues to run normally when this happens, so it seems the vehicle is not actually stopping.
Thanks again for your support — I look forward to your thoughts.
Hi @MatteoBucci -
Your tripod mounted base station seems to be tilted a bit back - you ideally will see the best performance in terms of reflections with it pointed at the horizon (level.)
I don’t think that fence would impact your 2.4 ghz WiFi signal much.
You say there aren’t WiFi networks nearby - but you’ve places a powerful WiFi transmitter right next to the Blue Boat radio! (The GPS.) I would guess that interference may be possible, but if you’re able to confirm they are operating on different channels with a phone WiFi scanner app that would be good. If they are on the same channel, that can lead to issues.
The mission download failure message happens when Cockpit times out waiting for a mission definition - this can happen easily with UDP packets going missing.
It sound like your omni directional antenna on the boat may be damaged. A VNA (vector network analyzer) can test these, or you could test with a replacement antenna.
Hi @tony-white ,
Thanks for the tips — next time I’ll repeat the test with the tripod leveled and the antenna pointing at the horizon.
The GNSS we’re using receives RTK corrections from a base on the lakeshore via UHF radio and is connected to the tablet via Bluetooth, so it should not be generating a Wi-Fi network. However, I’ll check with a Wi-Fi scanner app as you suggested, just to be sure.
About the possibly damaged omnidirectional antenna: can I swap it with the 4G antenna from the Cellular Modem Kit for testing, or do they have different specifications?
In case the BaseStation was powered on at some point without an antenna connected, how can I verify whether the radio has actually been damaged? If it were damaged, would I expect to see instability or disconnections with both antennas (omnidirectional and panel)? Or could it be possible that only one antenna shows issues while the other appears stable?
And in the event the radio is damaged, would I need to replace the entire BaseStation / GCS unit, or is it possible to replace only the internal radio module or specific parts?
Hi @MatteoBucci
The radio in the base station can be replaced individually. Poor performance is the only indicator it may have been damaged by not having antenna connected when powered.
Because the high gain antenna inherently improves signal, even if the radio is damaged you will still see better performance with it than the Omni. However if you’re not easily achieving 400+m with the flat panel, the radio is likely damaged.
If you’re not able to achieve 50m reliably with the Omni, the radio is likely damaged?
The antenna mast accessory greatly improves range by about 2x in most situations!
Hi @tony-white , thanks again for the help so far.
I did some controlled testing today in a office to verify the radio link behavior, and here are my findings:
As you can see in the attached photo, test setup:
With minimum Tx/Rx CCQ values around 98/75% and maximum 100/99.
Based on these measurements (strong RSSI, high SNR, and good throughput), do you think this could still indicate a problem with the omnidirectional antenna or possibly something damaged inside the GCS? (To me, these look like good link values).
At this point, I’m starting to think that the disconnections are more likely related to interference rather than hardware damage, as you had suggested earlier.
I noticed that the GNSS (Leica GS18) still generates a Wi-Fi access point when powered on, even if it’s not actively connected.
Since it was positioned close to the BaseStation, could this cause local interference or receiver desensitization?
I also saw that the radio frequency is set to “auto” in MikroTik. Do you think selecting a fixed 2.4 GHz channel (e.g., 1 / 6 / 11) instead of auto could improve stability?
I’m trying to understand whether channel selection and better RF separation could realistically explain the previous outdoor disconnections.
Hi @MatteoBucci -
While you are getting good reception, at that distance you likely would with any quality antenna!
The Leica may have been interfering, but AUTO means the radio is going to pick the best channel for you - likely automatically avoiding whatever the Leica is using. That said, RF energy an still “bleed” across channels and cause interference, especially at close range, so if you can turn that off it would be good.
A better test may be to separate the two radios by about 15-20, and (temporarily) attach the directional antenna too the BlueBoat! While the directional antenna is pointed straight up, try each of the omni-directional antennas on the BaseStation. You should see very similar results - if one antenna has significantly worse signal strength then the other, it likely needs to be replaced. Make sure to turn off radios when swapping antennas!
If they return similar results, the issue is therefore likely environmental, or due to a radio being damaged by operating without an attached antenna.
I ran some additional field tests today and wanted to share the results.
I first checked the Wi-Fi networks present in the area (see screenshot 1). The GNSS Wi-Fi appears as GS3626…, which should correspond to the Leica receiver.
I tested the link with the omni on the BlueBoat and the panel on the GCS, moving the boat to about 80 m. The lowest signal-to-noise ratio I observed was around 38 dB (screenshot 2).
I also repeated the same mission that previously caused the disconnections so I could compare the antennas (distant around 40m from the GCS):
Panel: never disconnected. The lowest signal to noise observed was about 42 dB (screenshot 3).
Omni: disconnected multiple times. When the disconnections happened, the signal to noise dropped to about 16 dB (screenshot 4).
Next Monday I will also try your suggestion of mounting the panel on the BlueBoat to see whether both omni antennas behave similarly.
Based on the data collected today, do these results look more like a hardware issue (for example a damaged antenna or radio), or more likely something related to the setup/environment (multipath or interference from nearby devices)?
Hi @MatteoBucci -
The critical piece of data is what range you’re having connection loss or the -16 dB signal to noise occur at? If it is close to 100m, that is expected for omni to omni, although ideal environments can increase this fairly dramatically. Adding a mast can about double the range performance. If you’re losing connection at 30m, then I’d suspect the antenna on either the BaseStation or BlueBoat is damaged.
