When folks talk about LoRa Mesh technologies, they're usually talking about either Meshtastic or MeshCore, the two competing protocols that have emerged and expanded like wildfire over the last couple of years. Mesh radio networks exist in most major areas, and there's quite an enthusiast crowd involved.
I recently posted a lot about Spectra, a LoRa Mesh Bridge that uses Veilid as the transport medium for LoRa nodes that aren't able to talk directly over radio due to distance, obstacles, etc. I was pretty excited about it, because it helps bridge gaps in these networks and let's folks connect when they might not otherwise be able to. Unfortunately, when I posted about the project in the MeshCore discord community I was flamed out of there relatively quickly with folks stomping all over the project as bad (though I did find one or two who genuinely wanted to learn more and discuss). The main concern, as far as I could tell, is that it would overload their mesh networks and so they were against it 100%.
This got me thinking... if a simple bridge that allowed ONLY text messages to flow across it would overwhelm their networks then... their networks aren't resilient at all. They're overly fragile and susceptible to jamming and DoS style attacks. But not only that... the experience left a permanent bad taste in my mouth with that community, and I already had a bad taste in my mouth about Meshtastic... so... I did what any enterprising hacker would do... I decided to just go and create my own protocol, with a focus on resiliency and throughput to ensure the network is as strong, stable, and fast as it can be. Utilizing techniques like TDMA and FHSS (Time Division Multiple Access and Frequency Hopping Spread Spectrum) you can reduce collisions, and be resistant to jamming and general RF noise.
And thus, SpectraMesh was born.
So, what is SpectraMesh?
SpectraMesh is a new LoRa Mesh Protocol, leveraging LoRa radios (like the SX1262) to communicate in the 915mhz frequency band (in the US/Canada, other bands coming in other regions). I looked at the weaknesses inherent in existing LoRa Mesh protocols like Meshtastic and Meshcore and attempted to address the shortcomings as best I could within the hardware limitations (it's half duplex hardware, so only so much can be done).
Key features:
-
TDMA - Time-Division Multiple Access is a mechanism to share limited RF bandwidth by assigning all radios a time slot to transmit in. Since everyone has their own time slot this dramatically reduces collisions in noisy networks. The drawback is it's a little inefficient with quieter or less busy networks. It can limit bandwidth some, but the lack of collisions in busy networks should smooth that out a lot.
-
FHSS - Frequency Hopping Spread Spectrum is a mechanism to combat noise and narrow band jamming by hopping between 130 different frequency channels. The radios all sync and hop frequencies together so they stay synced and able to talk while not relying on a single discrete frequency that could easily become noisy or jammed.
-
ADR - Adaptive Data Rates. As two radios start to talk to each other they negotiate radio settings that will yield the best throughput while maintaining a good connection. For shorter distances this will enable higher bandwidth and faster transmissions. For longer distances it will be slower and lower bandwidth - but they'll remain stable. Once done with the transmission the radios return to their previous states automatically.
-
AES-256 w/HMAC-SHA256 is used for network encryption and message integrity. ALL data on the network is encrypted with the network key, but additional private channels can be created with their own encryption. DMs are also privately encrypted between endpoints, using public key broadcast and TOFU methods to provide key trust. Trust, Verify and other key management commands are available.
-
Private channels can be created and shared with other contacts without requiring them to manually enter a PSK. a "/channel invite" command allows sending an special DM type (invite) that contains the channel information and PSK, and can be accepted with "/channel accept", which will then proceed to auto configure the channel for you.
-
GPS will be supported (manual entry of coordinates is possible, hardware support is present but I don't have the hardware to test yet).
-
Both TUI and Web companion apps are available - mobile will come soon (once I figure out how to do that).
-
Integration with Spectra Bridge starting in version v0.8.0 (coming soon)
-
KISS firmware for the Heltec V3 and Xiao ESP32-S3 WIO
-
Full UI firmware for the Lilygo T-Deck w/ support for all network features
More features will be added over time - expect this list to grow.
More details on SpectraMesh can be found on it's website: SpectraMesh
.