UE transmitting on its max power but still data failing

[RFSpecialist]

Also negative PHR is occurring when UE is at max tx power but UE still has data in buffer that it cannot send.

[Ryan]

@sadanandk2

Negative PHR values means-> When UE is already transmitting with max power and and gNB still instructing UE to transmit with more power through power control command.

Negative PHR help to anlyse cell edge scenarios.

Yup I agree!

Make sense but on other hands, max UE power is 23 dBm.

So if gNB tells UE to increase more 3 dB, let’s say so PHR is like -3 dB reported.

So like you know already UE reached its maximum 23 dBm power, why gNB tells him to increase more?

just curious really…

[sadanandk2]

gNBs job is to follow what scheduler needs.

This is how it works.

And importantly it help operator in knowing cell edge issues.

[ankgitm]

@Ryan

Assume UE reported power headroom 0, which means UE transmit on its max power 23 dBm.

Now let’s assume at gNB the signal (UL data packet) still failed and not successfully … typically on that case what will happen?

UE can’t increase more its power since it reached its maximum power, but still failure at cell side.

In this case it’s due to average UE distance more from cell.

You can fine tune pnom parameters…

Once it reach in negative value then impact seen in term of avg user throughput, low traffic on cell.

[Sakshama_Ghoslya]

@sadanandk2

gNBs job is to follow what scheduler needs.

This is how it works.

And importantly it help operator in knowing cell edge issues.

But the UEs max power is fixed and known to the scheduler.

Why gNB will ask UE to transmit with more power than its capacity?

Will that fall into UE’s PA’s non linearity reason causing damage to transmitted signal?

What are the max allowed transmit powers of 5G phones that we have?

[sadanandk2]

@Sakshama_Ghoslya

What are the max allowed transmit powers of 5G phones that we have?

Currently max is Power Class 2= 26 dbm for few high end devices.

@Sakshama_Ghoslya

But the UEs max power is fixed and known to the scheduler.

Why gNB will ask UE to transmit with more power than its capacity?

Will that fall into UE’s PA’s non linearity reason causing damage to transmitted signal?

I don’t know the exact reason, but I can think of something like this:

1. Even though UEs max power is fixed, but it is only known to cplane.
   I do not know when cplane would share it to mac scheduler.

2. Let’s assume somehow scheduler knows max UE power.
   But there are many scenarios like ENDC, CA where UE power is divided among the component carriers or among mcg / scg.
   And this power sharing is dynamic in nature.
   So scheduler may not be aware of exact power available with UE for that leg.

[Sakshama_Ghoslya]

This is informative. Thank you!

[Hamza_Bouchebbah]

@Ryan

I have one concern regard to UL data transmission.

Assume UE reported power headroom 0, which means UE transmit on its max power 23 dBm.

Now let’s assume at gNB the signal (UL data packet) still failed and not successfully … typically on that case what will happen?

UE can’t increase more its power since it reached its maximum power, but still failure at cell side.

Usually in such case UE won’t need to increase power but reduce the number of RB and use lower MCS.

It will continue until RLC SDU is bigger then scheduled TBS and MCS decrease will continue further until reaching MCS = 0

That is for sure as well will depends on vendor implementation as well I have seen some vendor avoiding MCS = 0 to avoid excessive segmentation.

[C_j]

@Ryan

UE can’t increase more its power since it reached its maximum power, but still failure at cell side.

Once UE reaches the maximum transmitting power, there is a tradeoff between Link-adaptation and Number of PRBs.

gNB will start reducing PRBs and reduce MCS as per Target SNR.

[moradiya]

@Hainm

UE report negative PHR mean UE can transmit over 23 dBm for a short time, right?

My understanding is slightly different…

If PHR is negative, means UE will use power according Pcmax.

If PHR is positive than will use calculated value as UE has still some power left.

[pappu_kumar_Chaurash]

@Hamza_Bouchebbah

Usually in such case UE won’t need to increase power but reduce the number of RB and use lower MCS.

It will continue until RLC SDU is bigger then scheduled TBS and MCS decrease will continue further until reaching MCS = 0

That is for sure as well will depends on vendor implementation as well I have seen some vendor avoiding MCS = 0 to avoid excessive segmentation.

How reduce in MCS will reduce UE calculated power in Uplink?

I had gone through power calculation expression given by 3GPP where variables are- number of RB for PUSCH, pathloss,TPC command.

[Hamza_Bouchebbah]

@pappu_kumar_Chaurash

How reduce in MCS will reduce UE calculated power in Uplink?

I had gone through power calculation expression given by 3GPP where variables are- number of RB for PUSCH, pathloss,TPC command.

1. No, MCS is not reduced to reduce power, but to maintain the BLER. As more power cannot be injected as you mentioned, as no headroom, so to enhance robustness LA reduce MCS.

2. Now how to transmit segment with limited power you segment the packet so power will be enough as if you segment packet less power required to transmit the same packet without sagmentation. Because one segment need less RB then full packet to transmit and with less RB you will need less power (power calculation depends on RB as well addionaly to tpc and all).

[sadanandk2]

@Hainm

UE report negative PHR mean UE can transmit over 23 dBm for a short time, right?

Negative PHR values means → When UE is already transmitting with max power and and gNB still instructing UE to transmit with more power through power control command.

Negative PHR help to analyse cell edge scenarios.

[pappu_kumar_Chaurash]

@Hamza_Bouchebbah

1. No, MCS is not reduced to reduce power, but to maintain the BLER. As more power cannot be injected as you mentioned, as no headroom, so to enhance robustness LA reduce MCS.
2. Now how to transmit segment with limited power you segment the packet so power will be enough as if you segment packet less power required to transmit the same packet without sagmentation. Because one segment need less RB then full packet to transmit and with less RB you will need less power (power calculation depends on RB as well addionaly to tpc and all).

Thanks! Here is what I understood. From both points.

1. Increase in Uplink power will decrease SNR in Uplink, so to keep low BLER gNB will reduce MCS.
2. Low size segments of RLC PDU will reflect low buffer on MAC. Will get less no of RB in UL grant which will indirectly allow UE to calculate less power.

Can we say from point 2: UL throughtput will decrease?

[Hamza_Bouchebbah]

Yes your statement is correct!

Yes, in point 2 throughput will decrease that is why I mentioned earlier.

Some vendor avoid MCS= 0 as it provide better robustness, but high segmentation and poorly reflected throughput on cell edge.