0
$\begingroup$

If you are in an environment with multiple transmitters of similar signal strengths, before band selection has taken place, how do receivers parse out overlapping OFDM signals from multiple transmitters?

Do systems just detect something off about the received packet and throw it out or is something else done?

$\endgroup$

2 Answers 2

1
$\begingroup$

If you are in an environment with multiple transmitters of similar signal strengths, before band selection has taken place, how do receivers parse out overlapping OFDM signals from multiple transmitters?

That's not a usual problem – OFDM systems you'll meet in practice either have fixed frequency allocations (e.g., DVB-T, DAB, Digital Radio Mondiale, or point-to-point microwave links), a central point of coordination that avoids such collisions (e.g., LTE) or use listen-before-talk and then back off-based strategies to repeat transmissions that were lost to collisions (e.g., Wifi, OcuSync).

Do systems just detect something off about the received packet and throw it out or is something else done?

That would be the last category, yes. You can detect things like multiple things that correlate like cyclic prefixes within one OFDM symbol's duration, or simply don't and let the packet be thrown out at the level where you validate checksums, or when the channel decoder fails to converge.

There's generally nothing stopping a receiver from employing successive interference cancellation, but typical OFDM use cases make that maximally hard: aside from the cyclic prefix, OFDM signals are for all practical purposes inherently well-whitened, and by design OFDM is being used in multipath environments. So, the rate loss that one would incur by making an OFDM transmission possible to decode in the presence of a similarly strong second one would be significant enough to make this infeasible.

$\endgroup$
2
  • $\begingroup$ Does it use a random time offset to resend if it doesn't get a reply? I'm curious if there is a single channel used for an initial connection. $\endgroup$ Jan 26 at 20:46
  • $\begingroup$ @FourierFlux that is extremely specific to the system, and with most of such systems, to the specific configuration. Think about how much larger you need to make the range of possible random relays to make it sufficiently likely that the "later" one doesn't collide again, if the distance between transmitters that collided might be 10 km vs if it can only be 2 m. $\endgroup$ Jan 26 at 21:20
0
$\begingroup$

I think frequency planning in case of cellular technology avoids frequency overlapping. But not quite sure in case multiple wifi ports, they may all transmit at 2.4 GHz or 5 GHz, and we may be detecting signals from all ports, but wifi connection is established through WPA authentication or so, until that time, overlapping signals are still detected I guess, and SSIDs are also detected as we can see in our terminals.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.