8

When describing 802.11b/g and the 2.4 GHz channels, or 802.11a/n with 5 GHz channels, textbooks (CWNA Official Study Guide) often show some figure with channels spaced such that certain channels cannot overlap and therefore ISI is avoided. The explanation is that ISI occurs in cases of multipath, where the signals of the same frequency interfere. This does make sense to me, i.e. that signals of the same frequency can interfere, and so non-overlapping channels would avoid ISI.

My understanding is that the 802.11 design assigns 1 signal per constant width of "frequency space" in a channel, for example

  • 1 signal per 2 MHz in a DSSS channel
  • 1 signal per subcarrier in OFDM channel

If we look at the figures in the URLs, the spacing of the channels and frequencies show a sequential order by Hz. What's not clear to me, is whether the same exact sequence must also hold for the timing of each signal's arrival. For example, consider 2.4 GHz channel 1: does the 2.410 GHz signal always arrive at the receiver before the 2.414 GHz signal? After the first 1 cycle, I would guess that the signals within a single channel arrive in order of lowest to highest Hz. But does this precise order hold after any length of time? What would prevent signals at different Hz from arriving at exactly the same instant? Or if simultaneous arrival of different-frequency signals is already expected to occur, then how does the radio discern which is the correct signal to receive?

1
  • I should have noticed that different frequencies implies different wavelengths. That may partially explain how the receiver can separate the different-Hz signals and process them without them interfering with each other. – T. Webster May 18 '13 at 8:15
7

These types of radio frequency transmissions are "frequency division multiplexed" (FDM). The different channels (ie, frequencies) transmit and receive at the same time. It's like a cocktail party where each conversation chooses a different pitch range -- the tenors have no problem hearing each other right through the basses at the next table.

In FDM, the receivers are tuned so that they only pickup transmissions in the channel (the frequency range) to which they are supposed to listen. So over a wider range (eg, the 2.4Ghz wifi) there are smaller divisions of frequency ranges into channels. The transmissions move at nearly the speed of light, so any transmitters on the same frequency -- the same channel in a give frequency band, eg 2.4Ghz's channel 11 -- will instantly be talking on top of each other. So any two 2.4Ghz channel 11 devices within range of each other will collide instantly.

Perhaps the part that isn't obvious is that electro-magnetic theory shows that superposition of E&M waves is not a problem. It's like waves in the open, ocean; You can have short spaced three foot waves (a highish frequency), moving over the surface at the same time you have long ocean swells (a lowish frequency). At the E&M receiver, you "simply" tune to the right frequency and you can pick the signals you want out of the noise. (And the E&M spectrums is very very noisy.)

2
  • I think your 3rd paragraph helped the most; what was most unclear to me was how the E&M receiver "simply" listens to the correct frequency, along with many others, simultaneously. – T. Webster May 17 '13 at 21:48
  • glad to be of assistance! – Craig Constantine May 17 '13 at 23:00
6

All this RF stuff sometimes makes my head spin, but I believe you are mixing up some concepts in your question. The separation of channels has absolutely nothing at all to do with ISI. ISI is an effect that occurs on the same frequency.

ISI typically results from multipath when more than one copy of a symbol arrives at the receiver at slightly different times and those copies start to overlap with the next transmitted symbol. This has a "blurring" effect, confusing the receiver and making it hard to understand the symbol.

The guard interval is what helps prevent ISI, by allowing the RF medium to "quiet down" before the next symbol is sent. A longer guard interval helps give the RF medium more time to become quiet, but reduces performance by reducing the amount of time that data is actually transmitted. A shorter guard interval increase performance but risks more chance of ISI.

A VERY loose analogy, think of someone talking into a microphone and that sound coming out of two speakers. If you introduce a delay in the transmission of sound from one speaker so that words start to overlap, this can make it difficult for people to understand what is being said. If that person were to give a full speech non-stop so the words are continually overlapping, it can be very hard to understand. However if they were to pause for a full second between each word, it would be easier to understand, but it would take longer to get through the speech.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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