So just to clarify, when you say perform IDFT you mean that we first add all the carrier signals in the frequency domain and then take it back to time domain to see the combination (something like that peak you draw). The diagram with carriers is in the time domain, that’s why I got confused.
Yes, the "carriers" are in the time domain. They are time domain waveforms at their given frequency. All real signals (ie. signals that exist in the real world) are time domain signals. I think you are confusing things when you say "add all the carrier signals in the frequency domain". If you are adding things in the frequency domain, then only things with the same frequency will add together (at each frequency). Signals at different frequencies don't "add together in the frequency domain". But two time domain signals at different frequencies can add together in the time domain. Hope this helps.
But didnt we use fixed point FFT and choose "power window" relative to noise floor? (FFT shifts) In that case we at risk of saturation inject "colour noise" to whole symbol, instead of compression problem
How would variable gain affect the OFDM system? I mean using compression enhancing the signal near the average and compressing the peaks. And doing the opposite at the receiver side.. ?
Good idea. And these sort of things are done in some systems. The problem is to design and implement the non-linear amplifier that would be needed. An alternate is to use non-uniform sampling levels, with more levels at the lower voltages. Other schemes are also used, which check for peaks in each symbol before transmission, and then rotate some subcarriers to reduce the peaks. For these schemes, it is necessary to set aside some data bits in the symbol in order to let the receiver know which subcarriers were rotated.
By “rotating carriers” you mean rotanting the phase to another constellation point on some carriers? Or to move a specific datapoint on the data stream to a different carrier according to some rule? (Like a scrambler) or is it something else? Seems like a good idea. And BTW Iain thank you for the video series, its a great primer on these topics Good work!
Thanks for the post... One query..in ofdma the subcarrier division is equal ryt...like say LTE each subcarrier will be 15KHz so this c1,C2,c3 waveforms u mentioned will not be applicable ..so how PAPR explains for ofdma? For this reason Uplink uses SC -Fdma ryt
Yes you're right - for the waveforms that I drew. I wanted to keep the diagram simple, so that I could explain the "phase alignment" problem, so I didn't draw waveforms for all the carriers. I probably should have mentioned it, but I only drew waveforms for the following multiples of the base frequency (ie. the 15 kHz frequency in the example you mention): 1, 2, 4, 8, 16. It was too complicated to draw all the other multiples (3, 5, 6, 7, 9, 10 ...). Also note that the diagrams I showed are "baseband". In other words, I didn't show them after they are up-converted to the carrier frequency (I'm pretty sure you realise that already). Thanks for your comment.
@@iain_explains beside PAPR factor, as i know that sensitive to carrier offset and drift also has influence on Ofdm signal, i hope in the future you will have a lecture like this to explain about that. It will be very helpful and i really appreciate what you've contributed, sir
you make what is difficult... very easy and fun to learn
That's so great to hear. I'm really glad you like the videos.
Your videos are pure delight to watch. Thanks a lot
Glad you like them!
I think PAPR mostly impact TX PA efficiency, that was a must aspect to explain
Thanks iain. J have never liked signals nd systems in my college. But your videos have made me fall in love with it.❤️
Glad you like them!
I'm glad that I found your channel ;)
Welcome aboard!
So just to clarify, when you say perform IDFT you mean that we first add all the carrier signals in the frequency domain and then take it back to time domain to see the combination (something like that peak you draw). The diagram with carriers is in the time domain, that’s why I got confused.
Yes, the "carriers" are in the time domain. They are time domain waveforms at their given frequency. All real signals (ie. signals that exist in the real world) are time domain signals. I think you are confusing things when you say "add all the carrier signals in the frequency domain". If you are adding things in the frequency domain, then only things with the same frequency will add together (at each frequency). Signals at different frequencies don't "add together in the frequency domain". But two time domain signals at different frequencies can add together in the time domain. Hope this helps.
Thanks for your sharing! But I have a problem. How does the PAPR affect the digital beamforming process? Looking forward to your detailed answer.
Excellent Explanation
But didnt we use fixed point FFT and choose "power window" relative to noise floor? (FFT shifts)
In that case we at risk of saturation inject "colour noise" to whole symbol, instead of compression problem
Thanks! So how to deal with that? Is CFR will be enought? How artifical reduction of the peak affect the signal?
legendary explanation
Thanks.
How would variable gain affect the OFDM system? I mean using compression enhancing the signal near the average and compressing the peaks. And doing the opposite at the receiver side.. ?
Good idea. And these sort of things are done in some systems. The problem is to design and implement the non-linear amplifier that would be needed. An alternate is to use non-uniform sampling levels, with more levels at the lower voltages. Other schemes are also used, which check for peaks in each symbol before transmission, and then rotate some subcarriers to reduce the peaks. For these schemes, it is necessary to set aside some data bits in the symbol in order to let the receiver know which subcarriers were rotated.
By “rotating carriers” you mean rotanting the phase to another constellation point on some carriers? Or to move a specific datapoint on the data stream to a different carrier according to some rule? (Like a scrambler) or is it something else? Seems like a good idea. And BTW Iain thank you for the video series, its a great primer on these topics Good work!
"Yes", to your first question. "No", to your second question. And thanks, I'm glad the videos are helpful.
Thanks for sharing may I know how did single carrier fdma solved this problem by adding n point dft block before the m point ifft where n
Thanks for the suggestion. I've added it to my "to do" list.
thank you so much
Good tutorial, thanks Iain !
Glad it was helpful!
Very nice and clear video! Thanks!
Glad it was helpful!
عشت ❤❤
please sir , could you talk about the frequency offset in a video ... ? thank you
Thanks for the suggestion. I've added it to my "to do" list.
مشكور ❤❤🎉
Greetings sir, can u plz teach us the "PAPR TECHNIQUES" le, PTS,SLM,CODING.... etc waiting for ur all further classes
Thanks for the suggestion. I've put it on my "to do" list.
Thanks for the post... One query..in ofdma the subcarrier division is equal ryt...like say LTE each subcarrier will be 15KHz so this c1,C2,c3 waveforms u mentioned will not be applicable ..so how PAPR explains for ofdma? For this reason Uplink uses SC -Fdma ryt
Yes you're right - for the waveforms that I drew. I wanted to keep the diagram simple, so that I could explain the "phase alignment" problem, so I didn't draw waveforms for all the carriers. I probably should have mentioned it, but I only drew waveforms for the following multiples of the base frequency (ie. the 15 kHz frequency in the example you mention): 1, 2, 4, 8, 16. It was too complicated to draw all the other multiples (3, 5, 6, 7, 9, 10 ...). Also note that the diagrams I showed are "baseband". In other words, I didn't show them after they are up-converted to the carrier frequency (I'm pretty sure you realise that already). Thanks for your comment.
Very helpful, Thank you.
Glad it was helpful!
Very useful !! Thank you, sir !
Glad it was helpful!
@@iain_explains beside PAPR factor, as i know that sensitive to carrier offset and drift also has influence on Ofdm signal, i hope in the future you will have a lecture like this to explain about that. It will be very helpful and i really appreciate what you've contributed, sir
Thanks for the suggestion, I've added it to my "to do" list.
Sir could you please give a real time example to reduce PAPR by using PTS. For better understanding purpose.
ma'am you have any pdf or lectures over PAPR TECHNIQUES
Thank you so much
You're most welcome