Sweet!
Another thought: I wonder if we could use an amateur band to re-encode the meshtastic signal and broadcast it on a BLOS capable frequency. Though, I suppose the AES nature of meshtastic would make it illegal to transmit on amateur bands. Unless the automated nature of it would fall under the classification of automated control signals?
Not going to happen. FCC makes those nearly all Military radio link reserved and also a couple of HAMs which you knowā¦ Snitch
So Palomar spans into the north valley, but the north side of that ridge is where we intended to place one relay.
Temecula Final 33.426667 -116.999167
Fair enough, worth a check.
Considering Iām still studying for my license, Iām just throwing out my āknow enough to be dangerousā thoughts here lol
I prefer transmitters not to be licensed. they require broadcasting a station ID. Look at lora in ham space or meshtactic on 433 mhz and the rules. We basically just need to find a way to make an autonomous relay not just repeater that resets the 7 hop max. It would allow us to deploy a much more closer together link of say 10 nodes and also would let us make tee off links. Thing is these relays would have to be smart. we would be up against making an algorithm for it to know the path to choose especially if we started adding multiple people beyond a tee and beyond the 7 hop max.
Does that make sense?
Its not unsolveable but would take studying and work
So; I dug around this morning in the meshtastic firmware and found this.
So the problem here is the data structure. If we could make 4 bits for the header, we could go up to 15 hops. That would make our network completely incompatible with other meshtastic systems though, and we would have to figure out how to separate our network from the other meshtastic systems.
Iām curious if the header has extra bits available.
More info:
HOP_LIMIT is stored in the last 3 bits of the header component uint8_t flags.
The other data stored in flags appears to only be want_ack, which is a bool. This means that if Iām correct, there are 3 bits available in the flags struct for us to play with.
Those extra 3 bits could theoretically give us 63 hops, but I donāt think we want to go that far.
The problem with this change is we need to move want_ack to get the extra bits, unless we fuck around with bit masking and put the first bit of hop max in the 4th place and leave want_ack in place.
Iām not even sure if that can work.
Our options are as follows:
0x07 to 0x0f and WANT_ACK_MASK from 0x08 to 0x10
0x07 to 0x17 and leave WANT_ACK_MASK alone.But the problem is I donāt remember if changing the hop mask that way will result in a decimal number that ignores the 4th bit. My C bit masking knowledge is a bit rusty.
Extra bits also pose another issue. The bigger you make it the more you need actual error correction to occur to make sure the header doesnt get messed up
Theres a lot to unpack here
Nope. No masking. Lets do it right if we are gonna mess with it. Good work on the reading of the firmware. Sorry I can only sporadically reply. No cell phone in the office and all
Something to chat about when we are home
This is doable and i think we would seperate from meshtastic purely because of different headers. we can tell it to ignore the header and message if location of those are wrong
They use bit masking on the flags var to extract the different flags.
There are 4 unused bits at the front of the 8 bit variable. The 5th bit is want_ack. The 6th through 8th bits are max_hops.
Just did the calculations, we canāt shove another bit in the 4th place and have the mask come out correctly.
The correct practice is to move want_ack.
Have you check if they are trully unused or if the data structure is expanded to handle some guard band information or other methods of error correction to correct the header?
Outside if that if this simple change works and we can reject others. Well we would then need to make a better node finding algorithm so that the least number of hops is reliably chosen in a dense area
Admittedly id love to look for potential algorithms for this
Fine by me
Ill pull the repos when I get home and start digging around. Its really limiting being in a secure facility
Bye for now lol
This method assembles the packet. The hop limit gets assembled with want_ack into the 8 bit uint. Still digging through the entire process, but it looks like itās just 4 empty bits.
Again, this will make our network completely incompatible with people not running our branch.
Yeah it would
We still have more physics to model. Like we are up against free space loss and weather.
Free space loss is something we can do to estimate the speed and config of the link. So heres the eq
Where:
d = Distance between the antennas.
f = Frequency
G sub t = The Gain of the Transmitting Antenna.
G sub r = The Gain of the Receiving Antenna.
c = Speed of light in vacuum ( Meters per Second)
We will need to know the expected power each location should see and thus that will determine our link config
Well shoot that did copy paste correctly. Fuckin office suites
RX boosted gain will help too. Antenna design plays a rolw due to vertical beam width even with omnidirectionals
I for sure dont want to climb if its going to fail haha
I might be able to do some less climbing intensive testing around San Diego. Hit up the coast guard monitoring system and ask if I can install the radio at point Loma, take the beneteau out to sea, ???, radio connectivity tests.
After next Sunday 17th getting up any SoCal mountain is going to get much more difficult.
Im not afraid of snow
My issue is long term deployment. Record snow yearsā¦ How to keep snow off the solar panels without heaters or wipers
I did a point loma simulation. It gets out to san clementeā¦ And it barely touches catalina
Would need repeater on san clemente which IS NOT HAPPENING UNDER ANY CIRCUMSTANCE 
If you wanted to fuck around with it on the boatā¦ Also just do it near the memorial park on loma. Near the highest point. Point the solar up and towards the south and stash it somewhere people wont stumble
It needs to be mounted on a non conductive pole
The paint and camo job needs to be as low observable as possible. It needs bits and hardware that isnt common. It needs tamper warnings and contains gps tracker etc. And a bro warning to all other tinkerers in the area
Cuz idk how unlicensed legal transmitters rules go when installing on pub lands when operating under 1w legally in an ism band
Equal effort. Coming out on the expeditions assuming this happens. Equal cost split of hardware. Help in the software dev efforts
We are just tickling the idea. For now its barely a twinkly dim star in the sky
My job involves me driving up and down California from San Bernardino to the Mexico border. I would be down to scout and do field work.
Company truck with company gas card!
Idk how much I could help with software dev but I am down for everything else.
The big relay inter connect is the hard part. I already have a local mesh in slc
But man connecting mountainous utah where I live and your pretty coastline is a steep task. Like the land is rugged and incredibly unforgivable haha
I didnt expect this much attention to what was initially just LOLs. I love this