2 nearby UEs use different RAT for Downlink and Uplink

Hi Experts.

While conducting 5G drive test for FTP Downlink and FTP Uplink…

Using 2 separate devices but in same car…

We find UE sticking to NR in DL handset but moving to LTE in UL device.

What can be the underlying cause?

Since time on NR KPI is being calculated for both DL and UL combined samples, this KPI is failing in many clusters.

Is there any chance of path imbalance in DL and UL?

And how do we check it and overcome?

Is there a parameter like buffer size by which traffic is routed to LTE leg or NR leg?

Is the BLER same for both DL or UL?

Coverage is always UL limited due to device power so if 5G is higher frequency than 4G then its likely device will fall back to 4G during high UL intensive test.

We are doing FTP with full buffer.

How can we contain UE to remain in NR?

Will PMAX change to higher value help?

Check UL split algorith threshols, UL SINR and pathloss.

Buffer size is only for splititing data to LTE also while NR ongoing.

UE is forced to rollback LTE only if UL SINR of 5G is low.

Can you please elaborate on UL Split algorithm threshold?

Are you hinting towards subframe configuration of NR TDD band?

Check this → Check NSA primary path (LTE or NR) from UE log

Sorry, I forgot to mention the network I’m working for is SA.

So how is your UL LTE and DL 5G SA?

LTE from vendor Samsung.

And 5G Ericsson for the network.

So one of the phones goes to LTE, not staying in 5G SA. IRAT HO.

That’s totally different issue.

Yes, correct.

UL handset is going to LTE via IRAT HO, DL staying back in NR.

Because UL and DL coverage is not same, that’s why.