Is Mesh FLARM Possible?

[RAS_Prime Administrator (OHM)]

You may be familiar with concept of "Mesh" WiFi.  Simply put it is a grouping of WiFi access points all tied together wirelessly to provide extended coverage.  For example placed around your home to help eliminate weak/dead spots.  This is similar in function to a WiFi "extender" but provides better access speeds.

https://www.tp-link.com/us/mesh-wifi/
https://en.wikipedia.org/wiki/Wireless_mesh_network

So what about using the mesh concept for FLARM?  PowerFlarm is said to provide up to 10km (5NM) of coverage [https://nadler.com/GliderPilotUSAflarmWeb/Flarm-WhatDoesItDo.html].

Example:  I am flying with four other FLARM equipped gliders.  Each glider is 5NM away from me in cardinal directions (north, south, east, west).  These gliders can see other gliders beyond my glider's FLARM's range.  If their FLARMs would rebroadcast what they see back to my FLARM then my range is effectively extended.  Add more gliders to the picture and my range gets every larger still.

There are limits to this rebroadcasting idea.  If there were many, many, gliders all rebroadcasting everything they see, plus everything that they are told about from even more distant gliders, then the system could quickly reach a point of wireless data saturation.  Then no one would get anything!  Plus I may not care about how a glider 200nm away is doing.

An idea to prevent saturation is to set a maximum radius of information rebroadcast.  Maybe my glider would only retransmit another glider's FLARM information that is within 10km from me (based on GPS position).

What are your thoughts?

[Dan Daly]

I get 15 km range, with lobes out to 40 km according to the CARP Range Analysis on flarm.com (image attached).  My transmit power out 40 km is over 5db according to KTrax range.  I have FLARM B, but I'm still 10km with only FLARM A.  I use a bottom-fed dipole 4.72" high as FLARM A, and my original FLARM antenna (shortened) as FLARM B. I keep about 10" from other transmitting antennae and GPS, and I keep my cellphone on airplane mode when I fly.  I don't use FLARM tactically, I'm happy if I see FLARM traffic reliably, and ADS-B and transponders are nice to have too.  Most gliders near me are FLARM equipped.

For me, MESH is meh.  I'd just improve my installation. Those old (my FLARM is from 2013)  FLARM-provided antennae are worth about $10US, and I'd replace them with something with good low-loss coax. The original coax is getting brittle (my opinion) which causes problems, as does tight bends and coiling the excess instead of shortening it. I have a new longer centre-fed dipole, and when someone complains about range, they go flying with the 'gold' antenna, which usually shows an improvement. They buy a new antenna, their problems go away. The FLARM is only as good as the weakest link...

My installation is in an SZD-55, which is fibreglass, which admittedly helps.

Dan

2D

[John Godfrey]

I would think that the much simpler approach is to opt into ADS-B. Much easier than reinventing it for Flarm.

[Moshe Braner]

In my version of SoftRF  https://github.com/moshe-braner/SoftRF I recently added a purposely-limited type of "mesh" capability.  It can optionally relay received data packets, with the following restrictions:

* Only one relayed hop in addition to the original hop.   The relayed packets are marked as such, and are not relayed again.

* Only packets from > 10km away are relayed - otherwise why bother?

* They are only relayed up to once per 15 seconds for a given source aircraft, and up to one relayed packet every 5 seconds in total (by the same relaying device).  That is, they take up to 10% of the available bandwidth - and that is subtracted from the normal transmissions of the same device.  Thus this does not add to congestion on the band at all, and also does not subtract significantly from the device's visibility to other devices, for collision avoidance.

* The settings allow choosing no relay, or limited relay of all traffic, or relay only landed aircraft.  The latter was the original idea (suggested by Caz Yokoyama).  Once a glider has landed (whether landed out, or back home) the transmission range is greatly reduced, thus a one-hop relay by the device in an airborne glider can help spread the message out to buddies and perhaps to an OGN ground station if within range of the airborne relaying device.

An issue arose, in that the relayed packet was received by the FLARM in the original sending aircraft, which then reported ITSELF as traffic.  That is confusing.  Too bad the FLARM software is not smart enough to ignore packets with the same ID as itself.  SoftRF does ignore such packets.  Also, who knows how the relay delay will affect the collision predictions done inside FLARM.  I therefore changed it so that the relayed packet is modified further in a way that makes it invisible to FLARM - partially defeating the "mesh" purpose.  The relayed packets from airborne traffic are thus only visible to SoftRF and OGNbase.  But packets from landed traffic are relayed unmodified, so they are still visible to FLARM and to OGN ground stations.  If you land out, keep your FLARM running, and if you then "see yourself" on your FLARM display, you are being relayed!

Comments and suggestions welcome.

On Sunday, October 9, 2022 at 11:35:07 AM UTC-4 RAS_Prime Administrator (OHM) wrote: