Handover Decision in 5G

Hi Experts,
It’s observed for 5G measurement report that UE reports cell level measurement along with SSB index measurement on which UE is latched.
Important point decision for Handover taken on basis of cell level measurement instead of beam level measurement.


I think it is the right way to decide HO.
When we say cell level measurement, these are nothing but beam level measurement averaged by some rules.
There are some parameters available for filtering.
Beam level measurement are average for number of (nrofSS-BlocksToAverage) beams, which satisfied condition of absThreshSS-BlocksConsolidation.
And this result is cell level measurement.
If none of the beam satisfy condition of absThreshSS-BlocksConsolidation, then highest beam measurement is considered as cell level measurement.
Same parameters are also for CSU RS measurement.

Yes indeed.
In some logs we see it that nrofSS-BlocksToAverage=8 and absThreshSS-BlocksConsolidation=-87dBm.

Threshold looks quit high.
So basically they want to use serving beam measurement as Cell measurement.

But isn’t this just for B1 event?
B1 event is like -120 dBm.
The threshold for B1 event.

These parameters are applicable for Intra NR mobility also.

If it is also for intrafreq measuremenst it makes sense to keep a high absThreshSS-BlocksConsolidation=-87dBm.
This means only one beam is involved in HO decision when RSRP is poor.
What relevance would have all the other beams?

Considering the fact that beam refinement process always ensures that UE latches to strongest beam always, HO based on single beam will delay the HO triggering.
If we consider other beams also, average would rather come down, and HO may be initiated earlier. Make sense?

I didn’t follow you: UE is handovering to a specific beam of the cell. What relevance will have to use absThreshSS-BlocksConsolidation=-110 dBm for example because UE would be force to average 8 SSB beams better than -110 dBm when only 1 is the good one.
It will delay handover decision like this ie it will trigger A3 too late.

Ok. You are thinking for target neighbor point of view, i am talking serving cell measurement point of view.
For example, if serving measurement is based on strongest beam (e.g -90 dBm), serving RSRP will be high compare to when serving measurement is based on 8 beams (avg -105 dBm).
So HO will be triggered earlier when we use 8 beam.
Similarly if we talk about neighbor cell measurement, if it is based on strong beam (nbr -85 dBm), ho will happen early as compare to when nbr is based on 8 beam (nbr = -100 dBm).
If number of beam to be measured at source and target are same, then there is no matter how many beams we are considering, HO boundary will be same.

Yes, you are right it works the same for both.
But with the setting I mentioned absThreshSS-BlocksConsolidation=-87dBm it means that only 1 beam is used to compare A3 for both source and target cell if RSRP of all beams is lower than -87 dBm.
So in weak RF all the rest of beams do not matter ,only teh strongest beam for source and target matters.
I mean measured beams by UE.

And if it is above -87 dBm things get complicated, because there could be 3-4 beams on source and 4-5 beams on taget cell fullfiling the criteria.
UE has to make the average for each serving and target cell, so it comes up with a single RSRP for serving cell and single RSRP of target cell and then compare for A3.

For 4G just need to check if RSRP based reselection and RSRQ based reselection there.
Then for high to low prio reselection.

Is it like “both RSRP and RSRQ” or like “RSRP or RSRQ”?

I configured snonintrasearch.
Snonintrasearch q r9,threshsrvlow, threshsrvlow r9.

Actualy there are many fading effects and rapid fluctuation if we consider single beam, specialy at higher frequencies.
So mobility based on single beam may increase number of handovers.
But this is totaly depend upon field we are working on.
Just my thought.

The setting I mentioned is for n78 3.5 GHz.

Reselection based on RSRQ is not supported in LTE.

Any idea what is cell range of 3.5G NR compare to 2.3G LTE coverage?
Just rough guess?

Oh, Thank You. I remember seeing parameter in SIB3 to configure, some feature there as well. :thinking:
Let me double check.

Huawei has limit to 14500 meters for 5G NR cellradius.
In 4G they have up to 200 km.

200 kms?
Or meters?

Yes, 200 km in 4G.