https://arstechnica.com/gadgets/2024/09/wavecore-blasts-gigabit-speed-signals-through-12-inch-thick-concrete/
Core drilling is tricky. Getting a 6 GHz signal through concrete is now easier.
Any idea how they achieved it?
Ultra directional antennas with maximum allowable power?
Which frequency? 5-6 Ghz?
I can't find it in the datasheet.
But it says power consumption of 36 W (very high for just L2 Ethernet bridge).
The article says that it's six gigahertz and that they are putting one unit on each side of the wall they want to blast through. So the signal is extremely strong right at the wall and then goes through it weakening and hits the other device and then is re-transmitted or whatever. By putting the routers directly on the wall, next to each other, they have enough power to get through the obstruction. But if they were in the middle of the room, for example, they probably would not.
Source: A passion for wireless technologies and myself being a licensed amateur radio operator.
It's the medium transfer that always kills you, this minimizes that by limiting it to 1 medium (concrete) and not doing air-concrete-air.
You can do eq and compensation for 1 medium, 2 media are harder.
2 media is concrete air or.... Its 3 media when you have add another air gap. And each transition is orders of magnitude of power requirments
Well, air-concrete-air is 2 media but 3 transitions. You can re-use the model for the transition which makes the math easier, but you still have to re-do all the math with different numbers, some of which you have to determine by analyzing the transformation the signal channel underwent through the transitions.
I don't think it would be an order of magnitude more power for EM trying to go through a transition between air and concrete, it's not like the wave is actually moving the matter so much as its bouncing into it sometimes, and the change in propagation velocity and group delay coupled with frequency rolloff in the transition, but it adds a ton of noise, mostly as the group delay gives you effective multipath, and especially modern modulation relies on heavily phase measurement (PSK and QAM) which can make everything all really icky when the phase interference is that small, we designed our modulation for large phase interference like multipath reflections, etc, the assumption that going between media would cause the reflected/refracted component to heavily rolloff, which might not be the case here.
Now, it's six gigahertz according to the article. But by putting the routers close to each other, it doesn't matter so much.