cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
3919
Views
44
Helpful
31
Replies

The Basics - How a Packet is transmitted over the RF Medium - Help Please

kfarrington
Level 3
Level 3

Hi Everyone,

Please could you help me understand a concept that may be basic to most of you, but for me, I dont fully understand.

When a packet is transmitted from a wireless NIC on a PC to an AP, does the copy of the same packet get transmitted many times in many directions by the wireless NIC or Antenna?.

Please can you be so kind and look at the picture I created.

Is this how it works.

Also, if this is correct, how many copies of the packet would the antenna send?

Many thx indeed to all for your help,

Kind regards,

Ken

7 Accepted Solutions

Accepted Solutions

Hi Ken,

Here's a helpful analogy that might be a bit overused, but I think it's still good :)

Imagine a lake with calm water. You pick a stone up and chuck it in, and it creates waves that radiate outward. This represents a very ideal situation for wireless, with no obstacles and no impedence, the waves continue outward pretty far, depending on the size of the stone (the radio power, if you will).

Now imagine you have an obstacle of some kind, let's say a large rock sticking above the surface of the water. If you now chuck a stone in, the waves radiate outward as before. However, when they hit the rock, what happens? You probably already know - the waves kind of bounce off the rock, but they also bend around it. In fact, aside from a small "shadow" directly behind the rock, the waves actually bend around and provide "coverage" behind it. In the end, you'll see those concentric circles extending far beyond the rock, again as long as there was enough power behind the stone dropping into the water.

When wireless waves hit an obstacle, there are difference, but the concepts are largely the same. Some waves will bounce off, and others will bend around it. Some will even go through the object if possible. This is why WiFi waves can extend throughout a building. Waves of this wavelength operate a bit differently than visible light. The color blue, for example, is about 450nm in length. It's not really possible for the color to penetrate opaque objects and refract around obstacles like a 12cm wave is able to do. You can't really think of WiFi waves like visible light.

Finally, that stone dropping in created quite a disturbance, but it was a single force that subsided pretty quickly. Now imagine a large pole that is pumping up and down in the middle of the lake. As it pumps up and down, it creates waves that radiate outward. If it pumps up high and down low, it can create large waves, but if it just barely moves up and down, it can create small ones. Regardless of the size of the wave that's created, it's the same wave that propagates everywhere.

Wireless is a push/pull kind of concept, where electricity is applied to an antenna and it pushes a wireless wave outward. The electricity is cut off and the wave cuts off as well. So whatever radiates outward will always be the same wave, until that wave hits a barrier or obstacle, as described above.

Sorry, that was a bit long. :( Does that help at all, and is that at all what you're looking for?

Jeff

View solution in original post

Ah, I follow now. I'll respond with a hastily-drawn MS Paint image :) Let me know if it makes sense. Basically, the 12cm is from any point in one cycle to the same point on the next cycle. So if you go from one peak to the next, or one valley to the next, that's the definition of a single wavelength, which is the significance of 12cm.

See also: http://en.wikipedia.org/wiki/Wavelength

And I don't think you want my brain... it's full of mostly-useless information, like how to beat Super Mario Bros. in under 10 minutes :)

Jeff

View solution in original post

Haha, no need to apologize, it's a confusing topic for sure.

You're almost there. The sine wave will actually go up and down 2.4 billion times (Mega = Million, Giga = Billion), and that's just how many times it goes up and down per second, not per wavelength. A "hertz" (Hz) actually has a unit of "per second", or 1/s, so it's describing how many times something happens per second.

12cm is the length of just one of those 2.4 billion waves. So the 2.4 billion waves extend quite far very quickly! Don't forget that these waves are traveling at the speed of light, which is darn fast. So in a single second's time, one wave is pushed out 2.4 billion x 12cm, or about 300,000,000 meters.

(Interestingly, and not really related to this, if you change the frequency you'll find that the wavelength also changes so that the wave will go the exact same distance in one second - 300,000,000 meters. That's because all light waves, regardless of frequency/wavelength, travels the same speed, which happens to be 300,000,000 meters/second. The calculation you used to derive 12cm, the 300/2400, is actually the formula c = wavelength*frequency, where c = the speed of light, which is again about 300,000,000m/s).

So to clarify my diagrams, think of the sine wave peaks as the black lines on my bottom drawing. Think of the valleys of the sine waves as being an invisible line exactly between the black lines. There is nothing significant about the black lines other than that they represent where the peak of the wave is. It has nothing to do with the thickness/length of the wave. As such, the wavelength is the distance between two black lines, or between two valleys. You can measure the wavelength from any point in the wave as long as you start/stop at the same point in the cycle.

You can't beat new video games all that quickly anymore, that's for sure. I did just beat Mario Galaxy last night on Wii though :) I'm a gamer nerd first, network nerd second, I'm afraid.

View solution in original post

What you're describing is part of the electromagnetic spectrum. See this: http://en.wikipedia.org/wiki/Electromagnetic_spectrum

Those designations are nothing more than titles for groups frequencies. When we talk about radio waves, microwaves, x-rays, gamma-rays, etc... all we're talking about is different frequencies.

Light can be as slow as 1Hz, which would be a 300,000,000 meter wavelength! It can also be as high (and higher) as 300,000,000,000,000 Hz, which would be a 1 micrometer wavelength. The spectrum is simply a range from 1 - infinity for frequencies that light waves can exist at.

So again, when you talk about different groups of frequencies, it's simply for the sake of organization. There's nothing different about any of them other than frequency and wavelength. The groups you list are all subcategories of the larger group called Radio Waves.

I can tell you want to learn. Can you tell I like to talk? :D This is neat stuff, because it goes far beyond the realm of WiFi.

View solution in original post

Hi Ken,

Wow, amazing pictures. Far better than I was able to make :)

I really don't like how text books show how waves get bigger as they expand. It is true that they get bigger, but they don't get stronger. Their amplitude doesn't increase, it actually decreases as it travels over free space. It makes sense, because signal strength decreases as we walk away from an access point or cell tower, and we grumble because we can't get coverage.

There's this concept of a "wave front" that describes, for like of better terms, the front or surface of the wave as it travels outward. This expands as it travels and does indeed get bigger, as they say. However, the amplitude gets smaller as the wave front expands.

Tired of analogies yet? Here's another one - Imagine a kid chewing some gum and starting to blow a bubble. As the bubble expands, the AMOUNT of gum doesn't increase, and yet it's getting bigger. The reason is that it's stretching out. The mass of gum continues to stretch thinner and thinner, causing its structural integrity to continue to deteriorate. So the gum is weaker at any given point despite the gum as a whole being bigger. Eventually, the thing pops and the kid has a mess.

For waves, there isn't a single mass of wave, per se, but there's a single amount of energy that the wave contains. This energy starts off very concentrated and "thick", but in a relatively confined space. As the wave travels outward, the surface or wave front continues to get bigger and bigger, but the energy is spread out over a larger area. So the amplitude is decreasing as the wave gets bigger.

The analogy breaks apart when talking about the bubble bursting. The wave never bursts, as we discussed earlier it simply fades to the background noise of the universe (assuming free-space travel out to infinity).

Strange, huh? As for your diagrams, that's the only thing I found to not be quite accurate. Very good and clearly-displayed info. It's clear you've done a lot of learning over the last few days. Major props for trying to figure this stuff out, I think a lot of wireless engineers don't care about understanding the physics behind the wireless signals. I'm with gstefanick, 5 stars for working so hard at this.

Jeff

View solution in original post

Hi Guys

Brilliant discussion and Jeff is one of the best guys you see here on wireless.

The last three pics you posted showing the isotropic radiator, this is the theoretically perfect radiator of which we only have the sun, ie radiating evenly in all directions.

From my degree days as you move from the source you use an inverse square law to degrade the signal strength as you move away from the source.

http://en.wikipedia.org/wiki/Inverse-square_law

I will be reading this thread more thouroughly and my view is that if you get your head around this you will be ahead of many engineers putting in WLANs. You need to be able to visualise what the radio is actually doing to do good planning and also see where the main pitfalls are.

There are some good explanations in the CWNA guide (sorry but it is relevant)

Just my two cents

View solution in original post

Ken, you're challenging us all to think about things that we often take for granted. Thanks for keeping us on our toes, I know that I've learned quite a bit myself by participating in this discussion!

To answer your first question, Ethernet over copper is not a frequency-based signal. Instead, electricity is used over copper to produce the signal. Generally, placing a positive voltage on the line produces a current in one direction, a binary 1, and placing a negative voltage on the line produces current the other direction, which is a binary 0. This is why Coax and UTP copper have two "cables", because a circuit must be completed between hosts. UTP has 4 pairs, and thus allows for 4 communication streams at once. This is why you can run full duplex using UTP, but with Coax you can only run half-duplex, since there's only one pair of copper (the inner-cable and outer-cable). Since the voltage applied is DC (as opposed to AC), there is no oscillation and no frequency.

I think your definition of a wireless carrier signal is very good. However, DSSS (802.11b) doesn't use a carrier wave, and OFDM (802.11a/g/n) uses carrier waves a bit differently (discussed below). My understanding is that the carrier wave is meant to be a heartbeat of sorts to wireless receivers, keeping them all calibrated and ready to receive more data. However, if an AP were to continually transmit a carrier wave, wireless clients would never be able to talk since the medium would always be busy. I assume (I wasn't able to verify this) that carrier waves can't be used when there are multiple speakers in an environment. But it would be perfect for radio stations and broadcast TV channels.

As for OFDM, all of the above still applies because it only uses carrier waves once it starts talking. Since an OFDM speaker communicates over a multitude of channels at once, each subchannel contains a carrier wave that is modulated to transmit data. This is because the subchannel must remain calibrated when it's not in use so it can be reused at any time. However, once the entire transmission is complete, the transmitter is silent. It doesn't use a "broad" carrier wave.

Something to consider about WiFi is that the start of any WiFi packet contains a series of alternating 0s and 1s for calibration. This is called the preamble. So this takes care of the fact that there is no carrier wave to keep everyone calibrated. Ethernet over copper also uses a preamble to calibrate everyone on the segment. For both WiFi and Ethernet copper, the medium is completely silent if there are no talkers.

I can't speak about fiber, I'm not so knowledgeable about the subject. As discussed earlier, Ethernet over copper doesn't use signals, so it doesn't use carrier waves.

Does that all make sense?

Jeff

View solution in original post

31 Replies 31

George Stefanick
VIP Alumni
VIP Alumni

Hello,

We can go as deep as you want 10101 or as high as you want ... you tell me ...

A quick overview.

A 802.11 frame has a lot of information. Source/Destination MACs, frame sequence, time stamp, SSID info, security info, and the list goes on...

A station will send a frame (from station to ap) after normal carrier sense is done, so not to have a collision with other frames (CSMA/CA).

In a 802.11 frame it has a sequence number. So when it is received at the AP, the AP knows the order. When a station sends a data frame to the AP, the AP responeds with an ACK frame. This tells the station the AP received its frame. If it doesnt get an ACK it will resend again.

I looked at your antenna pix. There is only 1 frame not multiple. Think of an antenna this way...

Suppose you and i are chatting. The foks around us can hear our conversation if we are loud enough. I dont need to repeat myself to the folks around us again(as you have in your pix).

Now lets add an antenna. Now i am screaming at you, and guess what the folks around us can hear me much louder now as well.

So, no frame 21 is not being reproduced it all directions. The only time frame 21 will be reproduced is if its not ACKd by the AP, assuming its an ACKable frame type.

I hope this help... please rate this post if it does ...

"Satisfaction does not come from knowing the solution, it comes from knowing why." - Rosalind Franklin
___________________________________________________________

Hi there, that is fantastic. Many many thx for the help. So the packet is sent only once from the Antenna.

The problem I am having is how does that 12cm radio wave go everywhere? I think it is the basics of radio waves I am having problems with.

I read somewhere that the wavelength size is 300/frequency in megahertz, 30 300/2437 is 0.12 meters

Im sorry if this sounds silly, but I am trying to visulise this data leaving the antenna and propagate everywhere in an omnidirection way.

Many thx for the help so far :)

Ken

Hi Ken,

Here's a helpful analogy that might be a bit overused, but I think it's still good :)

Imagine a lake with calm water. You pick a stone up and chuck it in, and it creates waves that radiate outward. This represents a very ideal situation for wireless, with no obstacles and no impedence, the waves continue outward pretty far, depending on the size of the stone (the radio power, if you will).

Now imagine you have an obstacle of some kind, let's say a large rock sticking above the surface of the water. If you now chuck a stone in, the waves radiate outward as before. However, when they hit the rock, what happens? You probably already know - the waves kind of bounce off the rock, but they also bend around it. In fact, aside from a small "shadow" directly behind the rock, the waves actually bend around and provide "coverage" behind it. In the end, you'll see those concentric circles extending far beyond the rock, again as long as there was enough power behind the stone dropping into the water.

When wireless waves hit an obstacle, there are difference, but the concepts are largely the same. Some waves will bounce off, and others will bend around it. Some will even go through the object if possible. This is why WiFi waves can extend throughout a building. Waves of this wavelength operate a bit differently than visible light. The color blue, for example, is about 450nm in length. It's not really possible for the color to penetrate opaque objects and refract around obstacles like a 12cm wave is able to do. You can't really think of WiFi waves like visible light.

Finally, that stone dropping in created quite a disturbance, but it was a single force that subsided pretty quickly. Now imagine a large pole that is pumping up and down in the middle of the lake. As it pumps up and down, it creates waves that radiate outward. If it pumps up high and down low, it can create large waves, but if it just barely moves up and down, it can create small ones. Regardless of the size of the wave that's created, it's the same wave that propagates everywhere.

Wireless is a push/pull kind of concept, where electricity is applied to an antenna and it pushes a wireless wave outward. The electricity is cut off and the wave cuts off as well. So whatever radiates outward will always be the same wave, until that wave hits a barrier or obstacle, as described above.

Sorry, that was a bit long. :( Does that help at all, and is that at all what you're looking for?

Jeff

Jeff, that is brilliant. Can I borrow your brain for a couple of years :) will hand it back after. promise!!

I get everything you have just said, but....

I still get stuck on the size of the wavelength thingy. 12 cm.

so are we saying that when you drop the stone, and it creates this wave (or wavelength) the wave is 12cm all the way round.

hang on, just doing quick picture.

Please see Lake Ken Pic :))

Thanks soooo much for the help with this mate.

Ken

Ah, I follow now. I'll respond with a hastily-drawn MS Paint image :) Let me know if it makes sense. Basically, the 12cm is from any point in one cycle to the same point on the next cycle. So if you go from one peak to the next, or one valley to the next, that's the definition of a single wavelength, which is the significance of 12cm.

See also: http://en.wikipedia.org/wiki/Wavelength

And I don't think you want my brain... it's full of mostly-useless information, like how to beat Super Mario Bros. in under 10 minutes :)

Jeff

10 minutes? I just got a PSP and can get past the 1 stage of 3d pacman. im useless at games

Im really sorry about this, but in one wavelength for 802.11g 2.4gHz, that sine wave on the top of your pic will go up and down 2.4 million times within the 12cm wavelength correct?

On the second pic on the bottom, the waves in a pond, are you saying that in between the lines is a full wavelength, ie 12cm, where as I made the actual line thickness the 12cm as the wavelength?

Im soo sorry mate, and I am trying to comprehend wifipedia now. some funky forulars in there?

Thx mate,

Ken

Also, I have just seen this.

Radio waves are divided into:-

Long Wave, around 1~2 km in wavelength. The radio station "Atlantic 252" broadcasts here.

Medium Wave, around 100m in wavelength, used by BBC Radio 5 and other "AM" stations.

VHF, which stands for "Very High Frequency" and has wavelengths of around 2m. This is where you find stereo "FM" radio stations, such as BBC Radio 1and Further up the VHF band are civilian aircraft and taxis.

UHF stands for "Ultra High Frequency", and has wavelengths of less than a metre. It's used for Police radio communications, television transmissions and military aircraft radios - although military communications are now mostly digital and encrypted.

and I am just trying to picture the pond scenario you describe?

You can tell I am not a Radio expert, but really want to learn.

Many thx once again,

Ken

What you're describing is part of the electromagnetic spectrum. See this: http://en.wikipedia.org/wiki/Electromagnetic_spectrum

Those designations are nothing more than titles for groups frequencies. When we talk about radio waves, microwaves, x-rays, gamma-rays, etc... all we're talking about is different frequencies.

Light can be as slow as 1Hz, which would be a 300,000,000 meter wavelength! It can also be as high (and higher) as 300,000,000,000,000 Hz, which would be a 1 micrometer wavelength. The spectrum is simply a range from 1 - infinity for frequencies that light waves can exist at.

So again, when you talk about different groups of frequencies, it's simply for the sake of organization. There's nothing different about any of them other than frequency and wavelength. The groups you list are all subcategories of the larger group called Radio Waves.

I can tell you want to learn. Can you tell I like to talk? :D This is neat stuff, because it goes far beyond the realm of WiFi.

Haha, no need to apologize, it's a confusing topic for sure.

You're almost there. The sine wave will actually go up and down 2.4 billion times (Mega = Million, Giga = Billion), and that's just how many times it goes up and down per second, not per wavelength. A "hertz" (Hz) actually has a unit of "per second", or 1/s, so it's describing how many times something happens per second.

12cm is the length of just one of those 2.4 billion waves. So the 2.4 billion waves extend quite far very quickly! Don't forget that these waves are traveling at the speed of light, which is darn fast. So in a single second's time, one wave is pushed out 2.4 billion x 12cm, or about 300,000,000 meters.

(Interestingly, and not really related to this, if you change the frequency you'll find that the wavelength also changes so that the wave will go the exact same distance in one second - 300,000,000 meters. That's because all light waves, regardless of frequency/wavelength, travels the same speed, which happens to be 300,000,000 meters/second. The calculation you used to derive 12cm, the 300/2400, is actually the formula c = wavelength*frequency, where c = the speed of light, which is again about 300,000,000m/s).

So to clarify my diagrams, think of the sine wave peaks as the black lines on my bottom drawing. Think of the valleys of the sine waves as being an invisible line exactly between the black lines. There is nothing significant about the black lines other than that they represent where the peak of the wave is. It has nothing to do with the thickness/length of the wave. As such, the wavelength is the distance between two black lines, or between two valleys. You can measure the wavelength from any point in the wave as long as you start/stop at the same point in the cycle.

You can't beat new video games all that quickly anymore, that's for sure. I did just beat Mario Galaxy last night on Wii though :) I'm a gamer nerd first, network nerd second, I'm afraid.

Jeff,

Thankyou soo much. This has clarified this for me. You have been a star and am going to rate all your posts here, as they desrve the highest marks.

Really hope we speak again soon mate, and keep up the gaming. Its the only thing that keeps us all sane :))

Once again, Many thx mate

Kind regards,

Ken

Thanks Ken! Always fun talking about this subject. Best of luck as you continue to study this abstract stuff :D

Oh, and the gaming will be kept up, for sure. Good luck beating that Pac-Man game, haha.

Jeff

:)

Hi Jeff (hope the gaming went well), and all :)

So have been looking into this further and there is an excellent tutorial on this to add to your excellent explanations.

But... I'm still not full there in visualising electromagnetic waves. MAN, This is soo hard.

I hope these links help someone else who is at the same stage as me in learning this stuff.

Very interesting view of how a dipole antenna electromagnetic wave looks like

http://www.colorado.edu/physics/2000/waves_particles/images/dipole.mov

And another interesting picture of an electromagnetic wave in motion

http://www.molphys.leidenuniv.nl/monos/smo/index.html?basics/light_anim.htm

Full Tutorial

http://www.colorado.edu/physics/2000/waves_particles/index.html

And visualising electromagnetic waves

http://www.colorado.edu/physics/2000/waves_particles/wpwaves5.html

Couple of questions if I may.

1. On the dipole movie, could someone explain this more in detail. There are three parts to this movie but I am not sure of their functions and relation.

a. The image of circles moving up and getting bigger on one side

b. The image of circles moving up and getting bigger on the other side, mirroring the first side

c. The white line that is apparent which connects the circles from both sides and has an arrow point in alternate directions.

2. On this computer generated image, this is a wavelength in motion. Few questions:

a. Is data only ever modulated onto the electric field and not the magnetic field

b. The wavelength shown here, (remember the 12cm question) this would be from one peak of the red to the next peak of the red correct (lets say length wise)? but how wide is it?

c. I have the word directional stored in my head. I still cant visualise this wavelength as an omnidirectional entity. (I am thinking that the antenna is sending this stream of wavelengths that are 12cm every wavelength, and the width of this wavelength is a certain value being pointed in one direction?

3. I have been reading that electromagnetic waves in space go on for infinity, in some other text. This is not correct? right?

Once again, sorry to ask these questions, but I still have a slight visualisation issue here :)

Thx all for the amazing help with this :)

Kind regards,

Ken

Hi Ken! Wow, you're really wanting to dive deep into the physics of EM Radiation. It's definitely a fun topic but incredibly abstract, and to be honest it's something I don't completely understand myself.

1. I can't view .mov files from my computer, so I'll need to let someone else answer these points.

2. While EM waves have two components, they aren't individually isolated components. As you can see by some of those nifty animations you found, the E and M waves oscillate in unison. So the wave is actually the entire thing; there aren't two waves that can be modulated individually.

As for your other questions, you're thinking of waves as a one-dimensional entity, and it's frankly hard to NOT do this because all our pictures of sine waves make them look this way. If a sine wave is a one-dimensional representation of a wave, then our analogy of the lake yesterday is a two-dimensional representation. If you take the cross-section of the lake, you'll see that the water forms a sine wave. Two-dimensionally, viewed from above the lake, this happens via concentric circles.

So let's consider a three-dimensional representation. (Have you ever played Final Fantasy X? If so you'll be able to picture this pretty well.) Imagine a huge sphere of water that's suspended in the air. In the center of the sphere is a balloon. It "pulsates" by expanding and contracting, and this pushes waves of water, in all directions at once, to the surface of the sphere. This is how waves work in three dimensions.

Now for the wireless client - imagine another balloon resting on the surface of the sphere. Now imagine that instead of "bobbing" up and down, this balloon expands and contracts like the balloon in the center, so that it's always barely touching the water's surface. As the waves travel outward, they cause that balloon to expand and contract at the exact same FREQUENCY as the balloon in the center of the sphere. The "data" has been replicated on the surface using the waves of the water as a medium.

I caps'd FREQUENCY for a reason there. The frequency of the pulses is perfectly replicated, but the POWER isn't. That balloon in the center is doing some heavy expansion and contraction, but by the time those waves reach the surface, they aren't quite as strong. The POWER of the waves (called the AMPLITUDE) is irrelevant to the "data". As long as the waves were strong enough to move that client balloon, that's all that matters.

So all of this is to answer #3. Waves have what we call "free space loss", which is how much the AMPLITUDE of the waves decreases over distance (without running into anything). The frequency never changes, just the AMPLITUDE. The EM Wave will go on for infinity in free space, but it will eventually fade to the background noise of the universe because the amplitude will become infinitesimal.

Going back to our water sphere, imagine that the sphere is only a few meters wide. We should have no problem seeing those balloons working fine. But now what if the sphere was miles and miles wide? If the water was 100% perfectly calm (and friction was non-existent), and our client balloon was perfectly sensitive to the waves, then it doesn't matter how big the sphere is, the data could be transmitted. However, the odds of the water being perfectly calm are pretty slim, and our balloon isn't going to be perfectly sensitive - it will miss tiny waves that don't have enough force to move it. So the waves will eventually be buried by the "background noise" of the disturbed water, and even if a tiny wave makes it to the surface, our client balloon might not even notice.

Whew, another long post! Let me know if you have any questions about any of this.

Jeff

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: