A radio system being "Multiple Access" means that more than
one transmission can occur in a geographical area. Virtually all wireless
communication systems are multiple access systems in one way or another
- after all it would be ridiculous if there could only be one television
station or FM radio station in each town.
The three main methods of achieving multiple access capability in a
communication system are as follows:
Frequency Division Multiple Access (FDMA):
This is the system that is used by commercial television and FM and
AM radio stations. On a television set, there are a certain number
of channels, and each channel carries one transmission. Each TV channel
corresponds to a certain frequency on which that channel is broadcast.
Since each channel is broadcast on a different frequency, it is relatively
easy for the television to "pick out" one transmission from
the others simply by tuning its receiving circuit to the desired frequency.
Another way to think of it is this: Imagine that there are several
pairs of people in a building who want to talk to each other. An FDMA
system is sort of like if each pair of people who want to talk go into
a different room from everybody else - then each pair will be able
to talk to each other just fine. The limitation here is on the number
of rooms in the building - there must be an empty room for each pair
of people. (In mobile phone systems, this means that there must be
an unused channel (frequency) for each conversation that needs to be
held.)
In a mobile phone system using FDMA, the mobile phone is assigned
a frequency and it only transmits on that frequency. If the phone travels
into a different cell than where the call began, it will be assigned
a new frequency in the new cell and will use that frequency instead
from then on.
Time Division Multiple Access (TDMA):
TDMA also divides the radio spectrum into channels like FDMA, but
it also divides each channel into "timeslices" of a given
length, then the transmitters assigned to each channel essentially "take
turns" transmitting, such that two transmitters on the same channel
never transmit at the same time.
In the talking analogy, a TDMA system would be one in which all the
pairs of people who want to talk to each other go into the same room
and simply take turns talking. The first pair of people might be allowed
to talk for 15 seconds, then the second pair can talk for 15 seconds,
then the third, fourth, etc... Once the cycle completes, the first
pair would then be allowed to talk for their 15 seconds, and so on.
Only in cell phone systems, the time slices are on the order of thousandths
of a second, not 15 seconds.
In mobile phone systems, TDMA is actually used in conjunction with
FDMA. The radio spectrum is first divided up into multiple channels
just like in an FDMA-only system, but then each channel is divided
into a certain number of timeslices. When in a conversation, a mobile
phone is assigned a certain channel and a certain timeslice. It then
only transmits on that channel and during that timeslice. As with FDMA-only
systems, if the phone moves into a new cell during the call, it is
assigned a new frequency and either the same or a new timeslice to
use.
How does the conversation "fit" into only one timeslice?
In TDMA systems, the voice from the phone's microphone is run through
a digital compression algorithm before being transmitted. This algorithm
is designed to remove portions of the speech sound that are not important
for human listening. With that information gone, the resulting signal
can "fit" into one timeslice instead of having to use the
channel all the time.
Why bother dividing the channels into timeslices? Why not
just use the digital compression and then make each channel smaller?
I don't know, but it seems like a good question.
Code Division Multiple Access (CDMA):
CDMA is the newest and most modern multiple access system that is
in use for commercial cell phone service. The way in which CDMA works
is actually quite technically complicated, but in essence, a CDMA phone
combines your voice with two specially designed pseudorandom codes.
Then, your phone transmits the resulting signal on the same channel
that the other phones in your area are transmitting on, at the same
time as the other phones are transmitting. However, the specific design
of the codes allow the computer at the base station to extract your
individual phone call from what would otherwise appear to be a massive
jumble of signals. (The channels that CDMA systems use are much larger
than the channels used by other systems - but this works out in the
end since you only need one channel in a CDMA system.)
The best metaphor I have heard for a CDMA system is this: Say you
have various pairs of people in a room who want to talk to each other,
and each pair of people speaks a different language (and no other languages).
So if you only speak English and you are talking to someone who speaks
English, since you can't understand what the people speaking in Spanish,
German, French, etc... are saying, you would easily be able to tune
out the sound of their voices and only listen to the other English
speaker. This of course is not exactly how CDMA works, but it can perhaps
be useful in understanding the system.
Steven Den Beste has written several good articles
about CDMA that you might be interested in reading. His articles
mainly concern actual cellular systems that use CDMA, but they also
cover the principles of CDMA itself.
|