Thanks @RFSpecialist.
Yeah, it’s DSS - In Ericsson it is Called ESS 5G NSA.
15 dbm is your SSB power and -110 dbm is B1 threshold, is it?
Yes. Again I am saying that RACH SR is not relevant.
One single UE placed at-110 dBm will destroy this KPI for the cell. 
That I got your point. 
But how i will insure in message NR PUCCH Power → 20 dBm?
In the DT log analysis you will see for UL measurements 3 UL measurements of power:
- PUSCH
- PUCCH
- RACH.
PUSCHis the first that gets capped at 20 dBm.
Then it comes PUCCH.
Then later, as path loss increases, you will get RLF, even RACH failure.
With UEs supporting dynamic power sharing, UE can go beyond 20dbm unless capped by the network.
What is UL path you set?
Is it over NR or NR and LTE or LTE only?
@Vicky, bottom point here is that you have a lot of users in 5G around -110 dBm.
Same like 4G, 5G link budget is uplink limited.
You may want to do a trial on some cells with B1 threshold -104 dBm for example, and then see how much RAHC SR will improve.
Also what you need to understand is that link budget during connected mode is different compared to link budget in idle mode (or at handover time when rach procedure is involved).
To be honest I am not so clear about link budget.
About path loss calculation of RACH = SSB block Power + B1 Threshold.
I need some time to study first and then we can chit-chat more technically and sensibly.
@Vicky, apart from Optimization changes to improve RACH SR, please also check what is RACH SR formula in your system.
Also there can be issue like Ghost RACH which reduces RACH SR.
Just additional thing which you can check.
This is totally understood. Reduction in B1 threshold will help to improve RACH to cut off the cell edge user.
I know and got your point.
We are using RACH for as per E\\ guideline
Acc_CBRA_RandomAccessSuccessRate = ( pmRadioRaCbSuccMsg3 / pmRadioRaCbAttMsg2 ) * 100
I am aware about Ghost RACH Issue and I know in Ericsson how we deal it.
Thanks for warning me again.
This low RACH SR we also obserevd in Samsung system, but for LTE.
There was issue at Model level which also consider Ghost RACH as a Attempt.
After correcting measurement formula and modem bug fixing, RACH SR improved from 10% to 60%.
Right, make sense.
We have internal system algo parameter to handle ghost problem by changing RACH detection power.
Anyway, thanks a lot.
By changing RACH detection power (by increasing RACH preamble target power in RCC on reconfig) you will eliminate ghost RACH, improve success rate but degrade user experience as more UEs will consume battery faster without reaching to gNodeB.
It may increase UL interference/RSSI, specially when RRC users are high in the cell.
But can give a try for this.
RACH consume in PRB is between 6 and 24 and lot of RACH ocassions so I don’t think will interfere each other much.
Depends on RACH settings indeed.
There is also one parameter like delta preamble message 3. You can increase that one too.
This is the delta increase in power between successful preamble and msg3 power.
Increase it by 1-2 dBs.
I have similar issue. Ghost UE is impacting NR RACH . Can you please guide which parameter i need to change to fix this
Could you pls tell how can we avoid this ghost RACH issue in Ericsson system ?
initial rach power should not be so high if it is -90 ish high chance of ghost preambles.