0:00

So as we said, the 2G standard of adoption process in Europe happen pretty

quickly. They kind of converse on GSM and TDMA

pretty fast, but in the US it was a little bit of a different story, a little

more complicated. So many companies in the US were familiar

with TDMA and they did prefer it but we'll look at another multiple access

technology right now. That was proposed and we'll actually

spend the majority of the rest of this lecture talking specifically about this

technology. So in 1988, this Cellular

Telecommunications Industry Association, or the CTIA for short, posted a set of

performance requirements that they wanted the new 2G standard to meet, and those

were requirements that the 2G standard had to have over analog.

One of those was a ten times improvement over the previous analog capacity.

So whatever digital standard came out, it needed to have a ten times improvement.

So, as we said, most companies were familiar with TDMA, but one in

particular, named Qualcomm [SOUND] had a different idea and that idea spawned from

their work in satellite communications and they used this other type of

technology. They said is, we have our typical

frequency and time. Rather than trying to fiddle out people

in this same diagram or the same space, either add more people and time, or add

more frequencies. Let's try to add another dimension to

this diagram, and then we'll separate users along that dimension instead.

So everyone can talk on all their frequencies and at the same time, but we

will separate them over a different dimension.

So you're probably saying, well, that's kind of weird, because what other

dimensions are there, that we haven't looked at?

Well, it comes back to our initial analogy when we pointed out the idea of

language, as we will see in a minute. Consider two transmitters and two

receivers, and this phone is trying to talk to this phone, and this phone is

trying to talk to this phone. So if they're transmitting at the same

time and a freq-, same frequency, we need another way to distinguish between them.

So what if there's a way that we can figuratively speaking lock each of these

links, so on this link has a lock and the only receiver that has a key is this one.

And on this link we also have a lock and the only receiver, that has the key, is

this receiver right here, so the idea is that even when this transmitter is

sending, and some of this power gets coupled into this receiver, that this

receiver doesn't have the key that's necessary, because it would need this

key, but it has this key, and that only will unlock this lock specifically.

So we call these different codes, and that's how we distinguish in what's

called code division multiple axis. So our third multiple axis technology is

code division or CDMA. So, each of these sessions or each of

these links, we have Anna talking, Ben, Charlie, and Dana.

Each of their lengths are going to have a separate code.

3:19

So, now we need what's called, a spreading code, okay, which is designed

by the engineers. A spreading code that's going to scramble

this message, and it's going to scramble this data signal.

So if you've ever heard of the process of encryption, that's similar to this and

what this is doing. It basically then, we multiply the data

signal by the spreading code, and we do that bit by bit.

And the spreading code, as you can see, changes much faster than the data signal,

so the spreading code is at a much higher frequency than the data signal is.

What happens then, is that it spreads the data signal spectrum.

And the frequency dome at this signal then is going to take up a much larger

range of frequencies and so that's why its called a spreading code.

4:09

So when we multiply these out let's think about what we get.

Well the first one is just 1 times 1, which is going to give us a one, and then

we have one times one again which is going to give us one.

Then we have one times negative one which will give us negative one and then one

times one again which gives us one. Then we have, then we have negative one.

So basically on the first end right here, we're just going to get the spreading

code back. So basically, we're just going to be

transmitting the spreading code because its data signal's a 1, and 1 times

anything is just whatever the other number is.

Now, for the negative 1, then we have to multiply the data signal by each of

these. So we have negative 1 times negative 1

gives us 1. It's going to basically invert the

spreading code on this side, so 1, 1, negative 1 minus one.

And this is what we actually transmit, we transmit the multiplication of the data

signal, and the spreading code. Then the receiver has a form of the

spreading code that they will invert, and then use to recover the original data

signal translated back down to its original form before it was spread out.

So this may seem easy in that you just give each receiver a separate spreading

code and it should work out. But it turns out that designing spreading

codes is actually a very complicated process, because the spreading codes have

to have a property, which is called being orthogonal.

Orthogonal really means that they will cancel out.

And so the idea is that if this receiver over here wants, is trying to receive

this message, that even when this person's message is going to come in,

that while His spreading code will be exactly the same as his transmitters,

which is what it needs in order to recover it.

The spreading code of this link, and the spreading code of this link, will cancel

each other out, so that when he applies his spreading code to this signal, it

will cancel out and basically give him nothing.

So he won't be able to recover any signal.

And that's exactly what you want, that's that self canceling property of the

orthogonal spreading codes, as we would say.

So designing spreading codes that have this property of being orthogonal is not

an easy task and that's what's really complicated about the idea of CDMA.

So now you're probably wondering, well why do we even bother using.

CDMA to begin with, and what did Qualcomm find so attractive about it?

It's just some complicated thing that maybe doesn't give us much improvement.

Well, it turns out that it actually does give many advantages.

The first is that the frequency reuse factor is now 1.

Do you remember the frequency reuse factor is the number of bands that we

have to use? Well, now, everyone's transmitting on

every single frequency, so we have no division there anymore, which makes that

factor Just one. So the frequency reuse is going to be

equal to one. Clearly that's desirable because that's

especially efficient in terms of the frequency sense.

Well the other thing that they found really, really enticing is that it was

expected to have like a forty times improvement.

Or with the existing analog capacity so greater then 4 times that 10 times

improvement that they that the CTIA was asking for for the next 2 use standard.

At least in their initial simulation so it came out to be 40 times what was

actually realized was not anywhere near as high as that but still 40 times is

very very enticing. So despite Qualcomm's claims, most

engineers at the time resisted the idea of code-based wireless.

First of all, it was a radical shift from the more intuitive TDMA type.

Secondly, Qualcomm really hadn't had a demonstration of CDMA on a large scale

yet, one that would show this 40 times improvement.

And so in 1989 the CTIA voted and approved TDMA as the first 2G digital

standard in the United States. But CDMA did have its share of problems,

and we'll look at those problems which had to be overcome before it did become

the cellular standard.