Signal modulation, how much data can be set?

[fredreload]

Hi Strange, can you tell me exactly how does digital radio demodulate signal into bits of 0s and 1s?

Edited August 18, 2016 by fredreload

[Strange]

Hi Strange, can you tell me exactly how does digital radio demodulate signal into bits of 0s and 1s?

 

 

Not in any detail, no. Partly because it is complicated and partly because it is different for different types of encoding and partly because I have forgotten most of what I knew about this!

[studiot]

Fred, I am guessing that your question was basically how to squeeze the maximum data rate down a given link.

 

I take my hat off to the ingenuity of engineers in the getting around the Shannon limit.

 

It should be noted that this limit refers to a single communication channel.

 

So any way of increasing the number of channels used to transmit the signal increases the data rate.

 

Here is an extract from Wikipedia (have you tried it ?)

 

 

In data communications a 56k modem will transmit a data rate of 56 kilobits per second (kbit/s) over a 4-kilohertz-wide telephone line (narrowband or voiceband). In the late 1980s, the Broadband Integrated Services Digital Network (B-ISDN) used the term to refer to a broad range of bit rates, independent of physical modulation details.^[3] The various forms of digital subscriber line (DSL) services are broadband in the sense that digital information is sent over multiple channels. Each channel is at higher frequency than the baseband voice channel, so it can support plain old telephone service on a single pair of wires at the same time.^[4] However, when that same line is converted to a non-loaded twisted-pair wire (no telephone filters), it becomes hundreds of kilohertz wide (broadband) and can carry up to 100 megabits per second using very-high-bitrate digital subscriber line (VDSL or VHDSL) techniques.^[5]

 

Note that this says for single voice channel (ie one telephone line) the limit is 56k, as I said earlier.

 

It also says that that single line is used as a multichannel line by a process I called frequency division multiplexing, earlier.

This is how modern megabit rates are achieved.

 

So that's one way to cheat the reaper.

 

Some years back, the BBC used another method when they transmitter digital code, they called Basicode.

 

This was very clever because it took an existing transmitted audio signal and superimposed the Basicode signal by chopping it up so the incoming audio signal looks (on an oscilloscope) rather like the audio I get from an old signal generator I have which uses the beat frequency technique to generate its output.

 

This will work on either the radio or the television audio.

[Strange]

Here is an extract from Wikipedia (have you tried it ?)

 

Link?

 

By the way, the link to 56K modems in that excerpt shows that the highest data rate achievable over standard telephone lines is the Shannon limit.

 

It was the switch to using those lines for digital data (e.g. removing filtering) that increased the bandwidth of the line and therefore increased data rates above 30 kbaud.

Edited August 18, 2016 by Strange

[fredreload]

Fred, I am guessing that your question was basically how to squeeze the maximum data rate down a given link.

 

I take my hat off to the ingenuity of engineers in the getting around the Shannon limit.

 

It should be noted that this limit refers to a single communication channel.

 

So any way of increasing the number of channels used to transmit the signal increases the data rate.

 

Here is an extract from Wikipedia (have you tried it ?)

 

 

Note that this says for single voice channel (ie one telephone line) the limit is 56k, as I said earlier.

 

It also says that that single line is used as a multichannel line by a process I called frequency division multiplexing, earlier.

This is how modern megabit rates are achieved.

 

So that's one way to cheat the reaper.

 

Some years back, the BBC used another method when they transmitter digital code, they called Basicode.

 

This was very clever because it took an existing transmitted audio signal and superimposed the Basicode signal by chopping it up so the incoming audio signal looks (on an oscilloscope) rather like the audio I get from an old signal generator I have which uses the beat frequency technique to generate its output.

 

This will work on either the radio or the television audio.

Hmm, I'm not sure about the encoding part but my idea originally is that for a single wavelength with frequency modulation to translate it back into a signal, then translate that signal into a value. That would be hard because you are reading in a signal, it would be much easier to decode these type of things inside a computer. Now if someone could answer how they keep the both sides in sync, do they need to be in sync? That would be cool

My new idea is that, you send out a beep after like a set amount of seconds within the limit constraint, for instance beep then 10 seconds then beep, I translate that 10 seconds into binary, yes I would need a converter to go from 10 to 1010 in binary, but my idea is that you decode the signal with software but not hardware, all that the wireless adapter does is read the signal. Now this 10 seconds could be in 10 milliseconds or less depending on the constraint. So that is how I transfer my data, let's say the time between the beep is 127 milliseconds beep-------------127ms------------beep, convert that to binary would be 01111111. So it consists of short pulses of time length for wireless signal, of course this is much easier to process in software

 

Basically like a remote, as Google suggested

 

Hmm, nevermind, that means time will always be a constraint. Well now I have to look up how to analyze signals

Also some suggests a factorial decode on frequency modulation, you know instead of 0s and 1s for radio signal you go 1234,1243,1423,...etc. For 4 bits you will get 2^4 which is 16 different combination, but for 4! = 24 different combination. Of course these all require a software look up to decode the actual value, I don't think you can decode factorial with hardware, I can be wrong

I take back what I said about universities

 

http://www.scienceforums.net/topic/97567-function-of-x-variables-with-non-repeat-outputs/

Edited August 18, 2016 by fredreload

[studiot]

Although you have quoted my entire post#28 I am struggling to see the relevance of any of your reply to it.

 

[fredreload]

Although you have quoted my entire post#28 I am struggling to see the relevance of any of your reply to it.

 

My bad, let me try to explain. Take a look at this. They use the example characters a, b, c, and d. Now this isn't one wavelength contrary to my post, but it is interesting and apply to the topic. Now imagine you send an RF signal of 4 wavelengths. _|-|_|--|_|---|_|----|_, I just increase each by 1 length but this 4 signals represents a,b,c, and d. You can place them in any order or have them repeats. Now how many data you can encode in these 4 signals? You can encode a value from 1 to 253 which is roughly 7 bits, with a digital signal you need 7 bits to complete this. Now you send this RF signal to another computer, I reads in 4 wavelengths, do an O(1) calculation to find the index and convert the data value to binary I dunno how many order the conversion would take but it shouldn't be that bad I think, instant data

 

P.S. Now you just need a transmitter and sensor and you can send data just about anywhere at the click of a button

Edited August 18, 2016 by fredreload

[Strange]

Hmm, I'm not sure about the encoding part but my idea originally is that for a single wavelength with frequency modulation to translate it back into a signal, then translate that signal into a value.

 

I'm not sure what this means - maybe because of your non-standard use of words.

 

The "single wavelength" that is modulated is called the carrier.

 

The thing that modulates the carrier is the signal. This is the information you want to transfer. It could be analog or digital, and encoded in any way.

 

 

That would be hard because you are reading in a signal, it would be much easier to decode these type of things inside a computer.

 

Much of the decoding of complex systems (e.g. GPS or mobile phones) is handled in software. Partly because it is easier, but also because it gives you the flexibility to handle changing or different protocols with the same hardware.

 

 

Now if someone could answer how they keep the both sides in sync, do they need to be in sync?

 

They synchronize by using information in the signal. For example, the information transmitted in "packets". Each packet has a unique bit-pattern at the start so it can be recognised as the start of a packet.

 

Note that this synchronization is not "absolute" but relative to the arrival of the signal. After all, receivers at different distances will receive the signal at different times.

 

 

My new idea is that, you send out a beep after like a set amount of seconds within the limit constraint, for instance beep then 10 seconds then beep, I translate that 10 seconds into binary, yes I would need a converter to go from 10 to 1010 in binary, but my idea is that you decode the signal with software but not hardware, all that the wireless adapter does is read the signal. Now this 10 seconds could be in 10 milliseconds or less depending on the constraint. So that is how I transfer my data, let's say the time between the beep is 127 milliseconds beep-------------127ms------------beep, convert that to binary would be 01111111. So it consists of short pulses of time length for wireless signal, of course this is much easier to process in software

 

There is an old joke that you can encode the entire contents of Wikipedia (for example) into a single mark on a stick. You measure the distance of the mark from the end of the stick. Convert that to binary and, as long as you measure it precisely enough, it can have any number of bits.

 

That is the problem with your scheme: how accurately can you divide up the time periods to decode the number being transmitted? Then you need to compare that with how many bits of normal (or compressed) data can be transmitted in the same time.

 

Measuring those intervals accurately would be much, much easier in hardware than software.

[fredreload]

 

I'm not sure what this means - maybe because of your non-standard use of words.

 

The "single wavelength" that is modulated is called the carrier.

 

The thing that modulates the carrier is the signal. This is the information you want to transfer. It could be analog or digital, and encoded in any way.

 

 

Much of the decoding of complex systems (e.g. GPS or mobile phones) is handled in software. Partly because it is easier, but also because it gives you the flexibility to handle changing or different protocols with the same hardware.

 

 

They synchronize by using information in the signal. For example, the information transmitted in "packets". Each packet has a unique bit-pattern at the start so it can be recognised as the start of a packet.

 

Note that this synchronization is not "absolute" but relative to the arrival of the signal. After all, receivers at different distances will receive the signal at different times.

 

 

There is an old joke that you can encode the entire contents of Wikipedia (for example) into a single mark on a stick. You measure the distance of the mark from the end of the stick. Convert that to binary and, as long as you measure it precisely enough, it can have any number of bits.

 

That is the problem with your scheme: how accurately can you divide up the time periods to decode the number being transmitted? Then you need to compare that with how many bits of normal (or compressed) data can be transmitted in the same time.

 

Measuring those intervals accurately would be much, much easier in hardware than software.

Hmm alright then, after the intervals are measured in hardware, can I have the intervals' lengths sent to the computer for further processing? How much time would that take?

[Strange]

Hmm alright then, after the intervals are measured in hardware, can I have the intervals' lengths sent to the computer for further processing?

 

Of course you could. But I can guarantee that this will be a much slower and more error prone than just transmitting digital data. And then the use of digital data means that you can do things like data compression and error correction, which allows for even higher data rates. Some very, very bright people have been thinking about this for many decades.

 

You need to learn about all the advanced techniques used (and how to evaluate and test them) before attempting to come up with something better.

[fredreload]

 

Of course you could. But I can guarantee that this will be a much slower and more error prone than just transmitting digital data. And then the use of digital data means that you can do things like data compression and error correction, which allows for even higher data rates. Some very, very bright people have been thinking about this for many decades.

 

You need to learn about all the advanced techniques used (and how to evaluate and test them) before attempting to come up with something better.

D:, did you come up with the factorial idea and tried to mask it from me?

[studiot]

 

Strange

 

I'm not sure what this means - maybe because of your non-standard use of words.

 

There is an old joke that you can encode the entire contents of Wikipedia (for example) into a single mark on a stick. You measure the distance of the mark from the end of the stick. Convert that to binary and, as long as you measure it precisely enough, it can have any number of bits.

That is the problem with your scheme: how accurately can you divide up the time periods to decode the number being transmitted? Then you need to compare that with how many bits of normal (or compressed) data can be transmitted in the same time.

 

Really pertinent comments, strange, +1

 

Fred, I don't know if you have done any digital electronics (back in the dim and distant I did a college project to design and construct a remote pc keyboard) but real world signals look nothing like your sketches.

 

One of the biggest issues is telling the receiver when the signal starts or stops, ie when a data pulse begins or ends.

What coding you use is far easier to deal with.

[fredreload]

So it's a draw?

 

P.S. Hmm, your idea seems better for single wavelength

Edited August 18, 2016 by fredreload

[Strange]

So it's a draw?

 

What is a draw?

 

 

P.S. Hmm, your idea seems better for single wavelength

 

What idea?

 

A single wavelength can transmit no information.

[fredreload]

 

What is a draw?

 

 

What idea?

 

A single wavelength can transmit no information.

Ya, nevermind, I just kind of like the idea to compete

The Fourier transform idea you posted right after me

As a friend pointed out my audio signal is restricted to cosine wave, and after that I still need to decode the wave

The factorial idea, did you come up with it?

You are Asian right? Are you Japanese or Chinese?

[Strange]

Who mentioned Fourier transforms? Or factorials?

 

 

As a friend pointed out my audio signal is restricted to cosine wave, and after that I still need to decode the wave

 

Nonsense.

[fredreload]

I got the idea from your Shannon sampling theorem and digital radio image. To attribute signal arrangements to a single function so you don't need a look up table to see where the bits are, it takes time and database to store the look up table. With that said a 5 letters arrangements can crunch up to 12 bits of binary data. It takes 5 signals, but it makes for a pretty good compression algorithm

[Strange]

I got the idea from your Shannon sampling theorem and digital radio image. To attribute signal arrangements to a single function so you don't need a look up table to see where the bits are, it takes time and database to store the look up table. With that said a 5 letters arrangements can crunch up to 12 bits of binary data. It takes 5 signals, but it makes for a pretty good compression algorithm

 

 

Why do you make random responses to simple questions? Your comments often have nothing to do with the question asked, or even the subject of the thread.

 

Who mentioned Fourier transforms? Or factorials?

[fredreload]

I did, isn't it easier to store binary numbers as decimals? I think I am just new to compression lol

 

P.S. No hard feelings, you can trust me

Edited August 18, 2016 by fredreload

[Carrock]

Fred, I am guessing that your question was basically how to squeeze the maximum data rate down a given link.

 

I take my hat off to the ingenuity of engineers in the getting around the Shannon limit.

 

It should be noted that this limit refers to a single communication channel.

 

So any way of increasing the number of channels used to transmit the signal increases the data rate.

 

Here is an extract from Wikipedia (have you tried it ?)

 

 

Note that this says for single voice channel (ie one telephone line) the limit is 56k, as I said earlier.

 

It also says that that single line is used as a multichannel line by a process I called frequency division multiplexing, earlier.

This is how modern megabit rates are achieved.

 

So that's one way to cheat the reaper.

You seem to be confusing the 'Shannon limit' with bandwidth efficiency i.e. Transmission Rate/Channel Bandwidth.

Or you're being disingenuous about the 'Shannon limit.'

 

 

Argument by Wiki

from your Wiki quote https://en.wikipedia.org/wiki/Broadband]

However, when that same line is converted to a non-loaded twisted-pair wire (no telephone filters), it becomes hundreds of kilohertz wide (broadband) and can carry up to 100 megabits per second using very-high-bitrate digital subscriber line (VDSL or VHDSL) techniques.

Breaks a 'Shannon limit' unless extreme techniques (far more expensive than just upgrading to fibre optic) are used to reduce noise.

 

From another Wiki https://en.wikipedia.org/wiki/Very-high-bit-rate_digital_subscriber_line

Second-generation systems (VDSL2; ITU-T G.993.2 approved in February 2006)[6] use frequencies of up to 30 MHz to provide data rates exceeding 100 Mbit/s simultaneously in both the upstream and downstream directions. The maximum available bit rate is achieved at a range of about 300 meters; performance degrades as the loop attenuation increases.

Plenty of data about VDSL standards in this Wiki.

 

 

 

Is this an example of what you mean by 'getting around the Shannon limit'?

To install broadband:

Remove the bandwidth limiting filters between broadband modem and exchange to increase line capacity from about 56Kb/s to (say)5056kB/s.

A 5056kB/s broadband channel would be nice but doesn't break the 'Shannon limit.'

So assign a 56kB/s channel (up to ~ 4kHz) for analogue phone.

Assign a 5000kB/s channel for broadband.

'Shannon limit' refers only to one channel ie the 56kB/s channel so the 5000kB/s channel breaks the 'Shannon limit.'

any way of increasing the number of channels used to transmit the signal increases the data rate.

So by increasing the number of channels the data rate has increased from 5056kB/s to, er, (56kB/s + 5000kB/s).

 

 

 

 

The concepts that double sideband suppressed carrier is impossible and that the 'Shannon limit' can be overcome by multiplexing are so contrary to current knowledge that they should be confined to speculations until you provide evidence to support them.

 

I notice your OP http://www.scienceforums.net/topic/97477-signal-modulation-how-much-data-can-be-set/#entry937903 has been voted up so presumably at least one person accepts the above concepts as mainstream science.

[Strange]

I did, isn't it easier to store binary numbers as decimals?

 

 

How do you store data as decimals?

 

Do you try and store 10 different voltages in each memory cell? And get each logic gate to handle 10 different voltages?

 

There are very good reasons why we don't do that.

[fredreload]

You win this one, reverse full adder

[Strange]

You win this one, reverse full adder

 

Please try and make sense.

[bimbo36]

i have always wondered what sort of circuits , or micro controllers were able to generate electromagnetic waves ??

 

 

[Strange]

i have always wondered what sort of circuits , or micro controllers were able to generate electromagnetic waves ??

 

 

All circuits generated electromagnetic waves, to some extent, as noise. If you want to generate electromagnetic waves then you will need an appropriate output circuit/device. For example, a radio transmitter for radio waves, an LED for light waves, a resistor for infra red, etc.

 

And please don't go back to your old habit of posting random irrelevant pictures in all your posts. It is very annoying and not at all helpful.