DSPRelated.com
Forums

Why low number of subcarriers is needed for high path loss scenarios​ ?

Started by cogwsn 4 years ago9 replieslatest reply 4 years ago141 views

In one of the papers related to 5G numerology, I read this statement: 

"Low number of subcarriers is needed for high path loss scenarios because higher number of subcarriers result in high PAPR values which is not a good condition for PA usage."

I am not able to figure out: Why low number of subcarriers is needed for high path loss scenarios 

Regards

Sumit 

[ - ]
Reply by kazJuly 20, 2020

Hi Sumit,

I think it is too obvious, you must be kidding, sorry.

Low subcarriers means you can use PA more efficiently and hence have better power level without saturation issues. But with DPD stuff(at gNB, not UE I assume) this would be less of an issue. So the mystery paper can argue for that.


[ - ]
Reply by cogwsnJuly 20, 2020

Hi Kaz, 

:D Well the second part of the sentence I understood but the first part was not obvious:  

"Why low number of subcarriers is needed for high path loss scenarios?"

Here is the paper(open access): https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8263598

Page-3, second column, second last para. 

Regards

Sumit 

[ - ]
Reply by SlartibartfastJuly 20, 2020

I'm with kaz, the answer was in your question.   As the number of carriers increases the PAPR increases, which decreases the average power available for the signal, hence greater pathloss/shorter range.


[ - ]
Reply by cogwsnJuly 20, 2020

Ok, I see. Yes, it was indeed a simple explanation. 

[ - ]
Reply by Detlef _AJuly 20, 2020

In a DSL szenario with many carriers and high PAPR you just let the amplifier clip, which results in error bursts. These errors are spread over time with the descrambler and eliminated by the Reed-Solomon error correction.

[ - ]
Reply by jimelectrJuly 20, 2020

Up to a point.  At around 16 to 32 carriers, irrespective of modulation type, the statistics start approaching a Gaussian distribution, and PAPR is about constant.  Or so I recall from my time at a now-defunct cellular basestation power amplifier manufacturer in the early 2000's.  Should be applicable to OFDM, which was just coming out back then.  I think the Gaussian approximation requires all carriers be at the same power or at least close.

[ - ]
Reply by thirugvasJuly 20, 2020

Hi Sumit,

Number of subcarriers has to be reduced for long distance communication. Because the total probability of error for multicarrier systems in general depends on the number of subcarriers.


Regards,

Thirunavukkarasu G

[ - ]
Reply by kazJuly 20, 2020

Hi Thirunavukkarasu,

Just to stimulate discussion. In that case I will send no subcarriers and get zero error. Apologies if it sounds sarcastic but it is not. I was once told by biology teacher if we eradicate all humans there will be no Malaria or Corona for today's mindset... it is true theoretically and is a hidden corner of implementation choices.