How Much Data Can a Single Wave Carry?

[kushl.guptaa]

Hi,

we friends were discussing about data speeds, bandwidth and frequency when this debatable question came in mind: How much data can a wave carry?

well, we all are aware that there is a bandwidth (consider gprs) which is capable of giving some speeds to download...when using on a mobile we are actually working on the radio waves with air as a medium. now, consider we separate just one wave-one single radio wave out of that bunch (is this possible in today's physics, i hope it is)...the question is-what is the data carrying capacity of that single wave?

Say i have one wave of 10KHz (it is having some energy given by E=hv)...how much data in bytes can this wave carry??give me the maximum limit of data irrespective of the transmission technology (any modulation technique whatsoever) being used.

Thanks for your help in clarifying my/our doubts.

 

[shawl]

Is this a Atomic, Solid State, or Comp. Physical problem?

data transfer rate = bitwidth * frequency / 8 ;
unit= bits/second

 

[kushl.guptaa]

Depends how you can contribute to the requested discussion...may be in terms of Atomic or Computational Physics...

 

[ZapperZ]

Depends how you can contribute to the requested discussion...may be in terms of Atomic or Computational Physics...

It's neither.

This thread has been moved.

Zz.

 

[elbeasto]

I am actually very interested in this now that you bring it up. At first I thought I knew the answer. So as I started typing the answer I noticed there were a few practical things that did not follow my logic.

I initially thought that bits traveled 'on' the wave so to speak like a kid in a rollercoaster seat. Except in this example, instead of the rollercoaster traveling along the tracks, it is the tracks traveling across some medium with the kid and seat fixated to the track. So my initial answer to your question was 1 bit per wave. So in a 10KHz frequency it would be 10k bit per second. I quickly confirmed in my head "Ya! that makes sense because a cat5 operates on 100MHz and it is 100Mbs cable". Then I got to thinking... "wait a second... what about cat5e... that is just supposed to be better metal. And a cat6 replacement still uses the same sending and receiving node so how could those 2 devices account for the increase in data transfer?" That was about the point everything crashed on me.

I had previously thought that a frequency meant literally the cycles per second which made me think that was how often it refreshed/gathered/overwrote/interpreted/etc... So in a case where we have 2 NIC on 2 separate computers, I thought each NIC operated on the same frequency and therefore they sent and received in sync because they were naturally built 'on-time'. But then I realized this isn't practical from a production standpoint.

So now I am very interested as well... How exactly does a bit travel across any medium whether it be wired or wireless.

 

[sophiecentaur]

I think you are basically asking about the information carrying capacity of 'a channel' here. The only limit to the actual information carrying capacity of a channel is the ratio of the signal energy to the noise energy on the channel. That sounds counterintuitive, I know, but if you take long enough over analysing a signal and look at as long a sample of the signal as you want then you can keep increasing the data rate. (Or, conversely, reducing the bandwidth)

Starting with an elementary 'textbook' picture of a binary digital signal (a 'boxcar' waveform of rectangular pulses), this will use a huge bandwidth and require that the noise level is less than half (roughly speaking) of the peak to peak signal voltage if you want to do a simple 'slice' in order to distinguish between a 0 and a 1. This is hideously wasteful of power and bandwidth. If you low pass filter this waveform, you will reduce the noise level but the waveform gets rounded off and, as you reduce the bandwidth further, the noise level drops more but the symbols start to flow into one another. A simple slice will not reveal 0 or 1 but you can recover the data if it follows 'known' rules and if you take long enough over analysing it. There is an ultimate limit to this process which is determined by the signal to noise ratio. Shannon's work on communication theory actually places a limit on this and it is way below what can be achieved in practice.

So the answer to your question is that you would need to specify more about your channel in order to be asking an answerable question. ? see what I mean?