Signal modulation, how much data can be set?

[fredreload]

So I took an interest into signal modulation and I have not look into it yet. Pretty much for a signal you are able to modulate it, but to what degree? For a single wavelength you have amplitude and to what degree can the signals be passed and encoded? Also for the regular wireless internet, how is the signal passed and how can it be improved. I want to jeep the most amount of data into a single wavelength and I want to know how to do that. I'll look more into it tomorrow. Strange if you are reading this you get a head start

[Strange]

https://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem

[Sensei]

For the best result (bandwidth), data should be already compressed with appropriate (depending on data type) compression algorithm.

Raw text can take f.e. 1 MB (million characters), but its zipped (.zip or .7zip) version could be taking 1%-10% of original size. It has many repeating strings, that easily can be compressed.

Some protocols support internal compression f.e. HTTP web browser is mentioning in HTTP request which compression algorithms it supports,

and replying HTTP server, is using this info, to compress data, and thus save bandwidth.

However, compressed data are, or can be vulnerable to lost in transmission.

Video or audio stream send compressed but not received entirely, will have damaged regions of image or noise instead of audio (happens every time I am watching my TV from satellite and there is heavy rain or snow),

but if the same would happen to f.e. compressed data application, that would be unrecoverable lost and require resent of data.

(TV decoder can't tell satellite "please send me data again", like it happens with regular two-way communicating Internet)

 

ps. I am wondering why are you putting it in computer section. It has more to do with Engineering or Physics, than Computers.

[fredreload]

https://documentation.meraki.com/MR/WiFi_Basics_and_Best_Practices/Wireless_Fundamentals%3A_Modulation

 

Right, my bad, I just thought this seems computer science ish. I'd favor Quadrature Amplitude Modulation (QAM), althought I'm not sure what varying degree of amplitude it could generate

 

P.S. Compression is a good idea, my idea is that transferring of data does not need to be in bit form, because all you do is transferring data, and once it's done you can just decode it, it can be in Arabic numbers or something, haven't thought it through

Edited August 16, 2016 by fredreload

[Strange]

my idea is that transferring of data does not need to be in bit form, because all you do is transferring data, and once it's done you can just decode it, it can be in Arabic numbers or something, haven't thought it through

 

You don't have to transmit as digital bits. You could use amplitude (or frequency) modulation where there are 10 different levels representing the values 0 to 10, for example. The problem is it becomes increasingly difficult to detect exactly what value is bign transmitted. It is much easier to detect "high" vs "low", which is why digital signals are so commonly used. But you can encode multiple bits per clock cycle (as QAM does).

 

Compression may help. But you have to take into account the time taken to compress and decompress data, if you want to transmit stuff in real time. Algorithms like ZIP require the entire data to be present and then compressed; you can't do it "on the fly" (there are other algorithms that are better suited to that - MPEG, MP3, etc).

[fredreload]

 

You don't have to transmit as digital bits. You could use amplitude (or frequency) modulation where there are 10 different levels representing the values 0 to 10, for example. The problem is it becomes increasingly difficult to detect exactly what value is bign transmitted. It is much easier to detect "high" vs "low", which is why digital signals are so commonly used. But you can encode multiple bits per clock cycle (as QAM does).

 

Compression may help. But you have to take into account the time taken to compress and decompress data, if you want to transmit stuff in real time. Algorithms like ZIP require the entire data to be present and then compressed; you can't do it "on the fly" (there are other algorithms that are better suited to that - MPEG, MP3, etc).

Something like radio wave and radio receiver except convert it to wireless internet

[Strange]

Yes. (I think.)

 

Wireless internet (and Wi-Fi and Bluetooth, and GPS) all use radio transmitters and receivers. They just differ in frequency, type of modulation, and the protocols used.

[fredreload]

Yes. (I think.)

 

Wireless internet (and Wi-Fi and Bluetooth, and GPS) all use radio transmitters and receivers. They just differ in frequency, type of modulation, and the protocols used.

Right, but they modulate the radio wave as digital signal, not as analogue right?

[Strange]

Right, but they modulate the radio wave as digital signal, not as analogue right?

 

Yes. (Although it might be more accurate to say that the modulation is analog but the data transmitted is digital.)

[fredreload]

 

Yes. (Although it might be more accurate to say that the modulation is analog but the data transmitted is digital.)

Right but imagine transferring data as an audio wave, a lot more data would go through right?

 

P.S. Same goes for ethernet, not sure how though

Edited August 16, 2016 by fredreload

[Strange]

Right but imagine transferring data as an audio wave, a lot more data would go through right?

 

Not sure why. We can digitally encode much more data to be transmitted in the same bandwidth than we can send as analog data. We can send multiple video calls down a telephone line that was designed for a single, limited quality audio signal.

 

We get hundreds of digital TV channels from the same transmitters that used to transmit just 5 analog channels. When they turned off the analog TV signals, they were able to massively increase the number of digital TV channels.

[fredreload]

Well but if you look at the PSK modulation where I got my ideas from in the link I posted. You see that every one wavelength contains 3 bits 000 to 111 from 8 different directions. But if you look at an audio wave in terms of amplitude, let's just say the amplitude can vary by 1000 levels for a sound that can be made, each wavelength would contain 2^9=512 < 1000, 9 bits for every wavelength and this is just an estimate for what analogue signal can do

[Strange]

Your analysis doesn't make much sense. And what sense it does make seems to contradict your point.

 

Your PSK example encodes 3x8 = 24 bits per cycle.

 

And even at a more realistic 16 bit DAC for the audio signal, PSK is still getting 50% more data in every sample.

Edited August 16, 2016 by Strange

[fredreload]

Your analysis doesn't make much sense. And what sense it does make seems to contradict your point.

 

Your PSK example encodes 3x8 = 24 bits per cycle.

 

And even at a more realistic 16 bit DAC for the audio signal, PSK is still getting 50% more data in every sample.

Well how about frequency modulation? Using frequency modulation to transfer data

Edited August 16, 2016 by fredreload

[Strange]

Well how about frequency modulation? Using frequency modulation to transfer data

 

It doesn't make any difference to the amount of data you can squeeze into a given bandwidth. (PSK is basically a form of frequency modulation).

 

The big advantage of frequency modulation is that you can have an absolute reference to measure the modulation against. In other words, you know what the carrier frequency is and so you can measure exactly how far the modulation has shifted the frequency.

 

For AM, you don't have any such absolute reference - amplitude is relative and depends on the signal strength, how far away you are, atmospheric conditions, quality of the receiver, etc. So then you have to work out an average amplitude and get a rough ide by how much the modulation makes it vary. But it might be varying for any number of other reasons.

 

That is why FM radio is so much better than AM.

[studiot]

It might be a good idea to come back down to Earth and look at a few basics of modulation.

 

1) For analog signals there is a maximum depth of modulation possible for AM since you cannot 'modulate' the carrier below zero. This is not a restriction suffered by some other types of modulation, particularly angle modulation, which includes frequency and phase modulation.

 

2) Shannon's criterion can be bypassed by multiplexing techniques, time division or frequency division are common. This is how modern broadband signals squeeze more than Shannon down the phone line and also how modern (TV etc) broadcasting works.

 

3) Digital techniques are inherently more accurate and reliable than analog at the current state of technology. The reverse was once true but this has reversed. It does not mean to say that it will not reverse again.

 

4) Analog techniques and amplitude modulation techniques refer to different processes. You can have analog or digital signals transmitted via amplitude modulation.

 

5) The oldest data transmission technique is neither analog or digital but is still the most accurate and reliable. That is pulsed continuous wave, for example morse code.

[fredreload]

 

It doesn't make any difference to the amount of data you can squeeze into a given bandwidth. (PSK is basically a form of frequency modulation).

 

The big advantage of frequency modulation is that you can have an absolute reference to measure the modulation against. In other words, you know what the carrier frequency is and so you can measure exactly how far the modulation has shifted the frequency.

 

For AM, you don't have any such absolute reference - amplitude is relative and depends on the signal strength, how far away you are, atmospheric conditions, quality of the receiver, etc. So then you have to work out an average amplitude and get a rough ide by how much the modulation makes it vary. But it might be varying for any number of other reasons.

 

That is why FM radio is so much better than AM.

FM radio is better, we call it a draw?

[Strange]

FM radio is better, we call it a draw?

 

 

Call what a draw? Digital radio is even better!

[fredreload]

 

 

Call what a draw? Digital radio is even better!

You use audio signal to decode the modulation back into a number form.

Well you don't transmit binary data, that is a waste, imagine this, but instead of just 0 and 1 you encode numbers from 0 to 100, that and with frequency encoding, I'm going to sleep = =, you are bad influence Strange, baaaaad

Yes, you can jeep a few hundred different noise into a single wavelength, and you can decode that signal back into numbers , looks like I win , FM radio only modulates 0 or 1 for each wavelength, an entire wavelength

This will remind me

Edited August 16, 2016 by fredreload

[Carrock]

It might be a good idea to come back down to Earth and look at a few basics of modulation.

 

1) For analog signals there is a maximum depth of modulation possible for AM since you cannot 'modulate' the carrier below zero. This is not a restriction suffered by some other types of modulation, particularly angle modulation, which includes frequency and phase modulation.

In AM you can 'modulate' the carrier below zero.

In single sideband suppressed carrier you are basically just frequency shifting the original modulation.

This was widely used in long distance analogue phone lines and in commercial and amateur short wave comms.

 

2) Shannon's criterion can be bypassed by multiplexing techniques, time division or frequency division are common. This is how modern broadband signals squeeze more than Shannon down the phone line and also how modern (TV etc) broadcasting works.

Reference please.

 

3) Digital techniques are inherently more accurate and reliable than analog at the current state of technology. The reverse was once true but this has reversed. It does not mean to say that it will not reverse again.

Depends what you want.

There's generally a sharp threshhold for digital; you go very quickly from 'accurate and reliable' to nothing at all.

High redundancy modulation such as SSB and CW generally has a very soft threshhold, very useful in the presence of heavy interference or noise.

4) Analog techniques and amplitude modulation techniques refer to different processes. You can have analogue or digital signals transmitted via amplitude modulation.

 

5) The oldest data transmission technique is neither analog or digital but is still the most accurate and reliable. That is pulsed continuous wave, for example morse code.

Morse code is the best for low bandwidth comms when used by experienced humans; not very good for machines.

[studiot]

In AM you can 'modulate' the carrier below zero.

In single sideband suppressed carrier you are basically just frequency shifting the original modulation.

This was widely used in long distance analogue phone lines and in commercial and amateur short wave comms.

 

A carrier is represented by

[math]y = A\cos (\theta )[/math]

If you modulate theta you get one of the types of angle modulation.

If you modulate A you get amplitude modulation.

But A cannot be negative so the depth AM is limited by the value of A.

Reference please.

 

The maximum telephony bit rate is 56 kilobits/s on standard lines.

This is why the original modems did not exceed this rate.

 

How do you think megabit rates are achieved in broadband?

By multiplexing as I said.

 

Read the tech manual for a good broadband modem.

Depends what you want.

There's generally a sharp threshhold for digital; you go very quickly from 'accurate and reliable' to nothing at all.

High redundancy modulation such as SSB and CW generally has a very soft threshhold, very useful in the presence of heavy interference or noise.

 

The threshold is true and was a considerable issue in the early days of digital equipment.

Modern digital equipment overcomes this to a large extent.

 

So that, for instance, tellurometers and early distomats used analog for the final digits in the measurement.

Modern ones are completely digital.

Morse code is the best for low bandwidth comms when used by experienced humans; not very good for machines.

 

That does not disagree with what I said.

When detection of any signal at all is unreliable, coding the simple on/off is more reliable and accurate than attempting to convey information during an on period, which may or may not be received.

 

 

Comments in red.

Edited August 16, 2016 by studiot

[Carrock]

In AM you can 'modulate' the carrier below zero.

In single sideband suppressed carrier you are basically just frequency shifting the original modulation.

This was widely used in long distance analogue phone lines and in commercial and amateur short wave comms.

 

 

A carrier is represented by

 

[latex] y = A\cos (\theta ) [/latex]

 

If you modulate theta you get one of the types of angle modulation.

If you modulate A you get amplitude modulation.

But A cannot be negative so the depth AM is limited by the value of A.

 

[latex]\cos (\theta )[/latex] can be negative so why not A?

An amplitude modulated carrier with [latex]A=1 + B\sin(\omega_{i}t)[/latex] ie [latex]A>=0[/latex] suitable for a simple diode envelope detector:

[latex] y = (1 + B\sin(\omega_{i}t))\sin(\omega_{c}t) [/latex]

where [latex]B <=1[/latex]

(normalising the carrier amplitude and only using sine wave modulation for clarity.)

Multiplying out:

[latex] y = (B/2)\cos(\omega_{c}-\omega_{i})t + \sin(\omega_{c}t) - (B/2)\cos(\omega_{c}+\omega_{i})t [/latex]

ie [latex]y = [/latex]lower sideband and carrier and upper sideband

 

A balanced modulator can be used to balance out the carrier ie:

[latex] y = B\sin(\omega_{i}t)\sin(\omega_{c}t) [/latex]

Multiplying out:

[latex] y = (B/2)\cos(\omega_{c}-\omega_{i})t - (B/2)\cos(\omega_{c}+\omega_{i})t [/latex]

 

As I mentioned in my previous post, the carrier and/or one of the sidebands need not be transmitted; they merely make tx and rx design easier, while wasting power and bandwidth.

 

Reference please.

 

 

The maximum telephony bit rate is 56 kilobits/s on standard lines.

 

This is why the original modems did not exceed this rate.

 

 

 

How do you think megabit rates are achieved in broadband?

 

By multiplexing as I said.

 

 

 

Read the tech manual for a good broadband modem.

My falsifiable understanding is that wired broadband uses only available bandwidth of the wiring between broadband modem and local exchange high bandwidth multiplexer.

 

2) Shannon's criterion can be bypassed by multiplexing techniques, time division or frequency division are common. This is how modern broadband signals squeeze more than Shannon down the phone line and also how modern (TV etc) broadcasting works.

I used to work on digital repeaters for digitised multiplexed phone calls using time, frequency and space multiplexing before optical fibre came in; bypassing 'Shannon's criterion' would have rendered my work on maximising bandwidth nugatory.

 

 

I largely agree with the rest.

Depends what you want.

There's generally a sharp threshhold for digital; you go very quickly from 'accurate and reliable' to nothing at all.

High redundancy modulation such as SSB and CW generally has a very soft threshhold, very useful in the presence of heavy interference or noise.

 

 

 

The threshold is true and was a considerable issue in the early days of digital equipment.

 

Modern digital equipment overcomes this to a large extent.

 

 

 

So that, for instance, tellurometers and early distomats used analog for the final digits in the measurement.

 

Modern ones are completely digital.

 

Morse code is the best for low bandwidth comms when used by experienced humans; not very good for machines.

 

 

 

That does not disagree with what I said.

 

When detection of any signal at all is unreliable, coding the simple on/off is more reliable and accurate than attempting to convey information during an on period, which may or may not be received.

 

 

 

 

 

Comments in red.

In future please respond to my posts directly rather than editing quotes attributed to me.

[fredreload]

I got my idea from Silicon Valley TV series lol, good show

 

P.S. The idea will attribute to them and is made public

Edited August 17, 2016 by fredreload

[fredreload]

Frequency modulation can also be applied for Ethernet right? But why do you need Ethernet when you got wireless?

[Strange]

Frequency modulation can also be applied for Ethernet right?

 

 

No. Ethernet specifies exactly how the signals are encoded on to the wires. If you started changing the frequency, then it wouldn't work.

 

 

 

But why do you need Ethernet when you got wireless?

 

Wired Ethernet is faster, cheaper and more secure than wireless.

 

(I should add, I don't know how Ethernet is implemented at the physical level. But I suspect I am about to find out ...)

Edited August 18, 2016 by Strange