6 Comments
User's avatar
Chuck Till's avatar

AREDN at 900 MHz could be a game-changer. Those of us who live in heavily forested, rolling terrain with suburban population densities find the higher frequencies impossible to use -- without putting an antenna at 100 feet AGL to get above the tree canopy, and then we run into city and county ordinances.

Expand full comment
Steve Stroh N8GNJ's avatar

Chuck - We're in violent agreement. I really hope that we can make 902-928 MHz work. I could even imagine using multiple units (such as 902-906 MHz and 924-928 MHz) to create multiple LANs.

Expand full comment
Alexander, DL4NO's avatar

FEC (forward error correction): FEC is a mechanism used near the physical transmission level to eliminate small random errors. But remember that most digital transmission systems have a very sharp threshold (typically less than 2 dB of SNR) where the error rate rises dramatically. So unless you add lots of redundancy to the transmitted data stream, FEC does not push your transmissions much deeper into the noise. But it increases usable speed of wireless transmissions on reasonable paths and reduces delays. Remember what TCP/IP does: If a data packet does not arrive intact, it must either be requested again from the sender (TCP, using large data buffers and therefore creating delays, often in the seconds range) or you use mechanisms to fill the gaps, for example in voice and video streams (UDP).

Expand full comment
Steve Stroh N8GNJ's avatar

Alexander - My interest in FEC in Amateur Radio goes back to the work done by John Langner WB2OSZ in Direwolf in implementing (the then mostly just theoretical) FX.25 in DIRE WOLF, and later IL2P. Both significantly improved the overall throughput of packet radio by proactively correcting single bit errors. There's a point of diminishing returns with varying levels of FEC.

Expand full comment
Alexander, DL4NO's avatar

AREDN on 900 MHz: I am skeptical on omnidirectional antennas because of multipath propagation. You must use a symbol rate that makes the duration of a symbol much longer than the time delay differences between the different transfer paths. Otherwise one symbol might overtake the other. Experience with New Packet Radio shows that many tests had to use directional antennas even on 430 MHz. The biggest advantage of 900 MHz compared to 2.3 or 5 GHz is the better transparency of many materials. Many AREDN experiments are failing because available geographical data do not see such soft obstacles: They were created by RADAR satellites from above.

Expand full comment
Steve Stroh N8GNJ's avatar

Alexander - We don't know if 900 MHz will work out for use reflashed with AREDN software. Thus my highlighting it for potential experimentation. My "sense" it that it will work in a lot of scenarios like small rural towns with a small number of Hams where there's no other practical way to build a high speed network that's Amateur Radio only.

Expand full comment