What makes WiFi faster at home than at a coffee shop? How does Google order its search results from the trillions of webpages on the Internet? Why does Verizon charge $15 for every GB of data we use? Is it really true that we are connected in six social steps or less? These are just a few of the many intriguing questions we can ask about the social and technical networks that form integral parts of our daily lives. This course is about exploring the answers, using a language that anyone can understand. We will focus on fundamental principles like “sharing is hard”, “crowds are wise”, and “network of networks” that have guided the design and sustainability of today’s networks, and summarize the theories behind everything from the social connections we make on platforms like Facebook to the technology upon which these websites run. Unlike other networking courses, the mathematics included here are no more complicated than adding and multiplying numbers. While mathematical details are necessary to fully specify the algorithms and systems we investigate, they are not required to understand the main ideas. We use illustrations, analogies, and anecdotes about networks as pedagogical tools in lieu of detailed equations. All the features of this course are available for free. It does not offer a certificate upon completion.
Interference
Loading...
来自 Princeton University 的课程
Networks Illustrated: Principles without Calculus
40 个评分
从本节课中
Random Access in Wifi Networks
In this lesson, we will investigate WiFi, another type of wireless network. Rather than having stringent power control algorithms as we saw for cellular, WiFi relies on "random access" methods to manage interference among users in the same location.
与讲师见面
Christopher BrintonLecturer
Electrical Engineering
Mung ChiangProfessor
Electrical Engineering
So even once your device goes through the three things that it needs to do to
access WiFi, then it has to deal with interference once it's actually on WiFi
and on the access point. And it shows in correct rate.
So, interference comes about the fact that frames, or as we call them frames
here. And we'll see in a future lecture when we
talk about the internet, specifically what Frames are but call frames are the
piece of digital data transmission for right now.
There's three possible outcomes of a collision.
So, if your frame collides with someone else's, there's three possible outcomes.
First is that both frames are lost, and that's clearly the worst case scenario,
is that both then would have to be retransmitted.
It's just lost data. Second is capture, and what that means is
that one out of the two you would assume that it would be the one with the higher
signal to interference ratio. So the frame with the higher signal to
interference ratio will be kept, the other one will be lost.
So we basically, with capture, we keep one of the two.
The better of the two, one with better channel quality.
And third is double capture and in this case both of them are kept, even though
they collide they both get to the receiver and they're both received
properly. So, that's two out of two.
So, here we get zero out of two of them. Here we get one out of two and in double
capture, we get two out of two. Now which outcomes going to prevail
depends on a lot of different things. First of all the duration of the overlap,
so if they're overlapping for a long time, the longer they're overlapping the
more more damage it's going to do to the frames.
And so we would go more so towards having both of them lost the longer the duration
Second is the difference in the signal to interference ratios between the devices.
So if one device requires a much, or if one, if one of the frames has a much
higher signal-to-interference ratio, than the other frame.
Then we would expect probably a capture scenario would happen.
If the other one had a very Much smaller in a very low single interference ratio,
much lower channel quality. And a third is the required SIR at the
reciever. So if the receiver is requiring a very
high single interference ratio, clearly this collision is going to cause a lot of
inteference. So we'd be more so along the lines of
both of the frames being lost, but if the receiver's very good, and it can deal
with a lot of interference and still decode the incoming signal, the incoming
frame, than we might come more down to the double capture scenario.
So, this is just an example to look at here Let's consider two devices A and B.
Before we said the access point was A but here now the devices we'll refer to them
as letters A and B. A has an SIR of three and B has an SIR of
two to the access point. So, if they're both transmitting at the
same time, they both sent a frame out and this is their transmission ranges here.
Again, in circles which it only meant transmission range.
Both of them send data out at the same time.
They're going to collide. Before by tagging we will just look at
the collision happening at the receiver itself.
And so if this was the overlap here, for instance, we see that A's signal
interference ratio is up at 3 and B's is up at 2.
So if only one of them was going to survive this collision, we would expect
that it would be A because A has a higher SIR.
but that will also depend upon how much of the overlap we have.
So if the overlap is very long or it's a long time, then it's a much more likely
case that both of them will be lost or at least the B will be lost and the required
SIR so this, B access point here might have some requirement.
So, if the access point requires signal to interference ratio of three at minimum
then we know that two would clearly not make it.
But three also wouldn't make it because this is going to ruin the SIR of A.
But if it was down somewhere like, you know, 1 or somewhere between 1 and 0, it
might be okay. You don't really know.
The key here is that it's hard to tell. So it's hard to tell which scenario we're
going to fall in. It takes a lot of physics to be able to
model that, which is beyond Are prerequisites here, and so we're going to
assume always, that both of the frames are lost.
We'll just take a conservative approach, and that's a typical engineering
standard, is to assume the worst case scenario, and here we will assume that
both of the frames are always lost, whenever we have a collision.
So, next thing you're probably wondering is, well, to deal with interference, why
don't we just use power control? We spent a whole lecture on distributing
power control. See if it worked really nicely for
cellular why can't we just do that here? Well, there's a few things about WiFi
that you have to take into consideration that make power control not a very good
solution for it. If you just look at the difference the
typologies are even though they're you know, we can draw sa, correlation between
a cell tower, a base station access point.
The devices might be the same. And so on, and so forth.
There is a lot of differences in the, how the devices operate, and how the
communication occurs. So, the first is that WiFi is all in
unlicensed band. So it would be really hard for us to be
able to regulate the power control among all the devices, and make sure they're
all conforming to the exact same standard.
Second is that In cellu, in WiFi, we have a much smaller cell size.
I mean it's not really a cell. It's a basic service set here.
But BSS is typically less than a hundred meters.
It would have to be or else you wouldn't be able to get much reception at the the
edge of the BSS. Cellular can be greater than 1 kilometer
from the tower. Since orders of, it seems like at least
an order of magnitude difference there. And the range which means that you have
more devices and then the idea that is that with one device comes into.
And then, so wants to talk to this play station wants to communicate with it its
not really going to effect that distributed power control equilibrium
much, so we saw that the DBC algorithm would converge to some given parallels
whatever they were, and we'll just call this a midpoint.
This is not zero, just if you just maybe converse to some given levels, right?
And so then, if another device comes into the network, it may jump a little bit,
but it's not going to jump much and it's going to re-converge relatively quickly,
right and it'll come back actually clearer.
In WiFi if we were doing TPC, right? We would eventually get to the conversion
points and we get to the conversion points fast because there's a small
number of devices. But then if another device came into the
access point there's such a small number of them and on the side of them we should
draw you know, many, many, many more lines.
[NOISE] There's many more in reality. if there's only three devices in here and
another one comes, they're going to jump very rapidly because any one given device
coming in could have a big effect on that equilibrium.
As we can see, very different amount of channel [INAUDIBLE] for this short period
of time just as an example. So it could be that cell size will have a
lot to do with it. Third is that WiFi has a maximum transmit
power. It's much smaller because of the fact
that we're in an unlicensed band than WiFi, it has a much smaller maximum
transmit power. So, the maximum transmit power of WiFi,
you might say is somewhere around a hundred milliwatts, on average.
In cellular you could go up to 2 watts. So, you have a more dynamic range to do
this kind of a power control. To get to the levels that you need to be
if there's a lot of interference. Under Wi-fi you can't go above 100
millawatts because it's regulated. And for that sense we may not be able to
even come to a set of power levels that would work for us.
And fourth is that Wi-fi is typically indoors.
And that means that there's a lot more that we might bump into.
There's a lot of walls that can get in our way.
A lot more interference that could go on inside of a house than outside.
When you're talking the cell tower. The cell tower's also, much more
elevated, so it's easier to have a clearer channel to the tower itself.
