Yes, for 5G mmWave n257G (4).
Only for that one.
But this doesn’t prove that each antenna in uplink is directional.
The website is very good, but it has a lot of missing information.
It does not list all UL MIMO streams number for UL bands.
UE has a lot of antennas for different bands.
I didn’t say that each antenna in the UE will be directional, I only stated that I think FR2 antennas at the UE are directional.
This link has a view about different antennas at the UE, I think this link is very useful to look at:
Yes, please read this from that site:
It is about RX beam
cacombos site does not show errors, it is mapped fro the enitre UE capability message.
So no band is missing or combinations for that country.
But indeed it is missing for all countries.
So in some countries some band combinations might be missing/present based on country.
I am also confused about this: “Antenna for mmWave bands is intrinsically directional”
Yes, I was only talking about the missing number of UL MIMO streams in some UL bands.
Other than that it is a really good website.
This make sense, especially for FR2.
UE is always considered with front side UP, sitting on the table.
But you can have it on your hand in which case the semi-sphere concept is not working well.
With a semi-sphere directional antenna you easily end up into no communication with gnodeB because you are on the backside of the semisphere.
Hence there will be no communication between UE and gNodeB,
No matter what phase shifts you apply,
This is why beam sweeping is used at the UE side to cover the whole 360 degree space around the UE and catch SSB beam from any possible direction 
All these pictures are proof of concept and are again for downlink only.
I am speaking about Uplink.
The idea is that in uplink in the Ue there is no massive MIMO to compensate link budget of Downlink.
Link budget is always Uplink limited.
And also there is no uplink beamforming in the UE.
We have to make a clear difference between UL MIMO and beamforming in UE in Uplink.
After gNB and UE establish a beam pair (suitable Uplink beam with suitable Downlink beam) they can communicate in both directions UL and DL.
Of course, Cell radius in the end will be limited by UL.
I’m not sure about that, so I can’t comment.
These pictures are not proof of concept, this is the beam management procedure defined by 3GPP and consist of 3 phases P1, P2, P3 and I think the answer to your questions can be found by understanding these procedures in details.
I am aware of the Downlink procedure, I understand them.
What I am trying to say is that number of antennas for Downlink and for Uplink are not the same.
This is why there is no Uplink massive MIMO in the UE.
This would work only for TDD and only if there are same number of antenna for UL and DL.
You are correct, Massive MIMO in the Downlink is composed of many arrays, each array is capable of communicating with UE according to UE capability.
For example, if the UE supports 2 ports, gNB will us use 2 ports to communicate with the UE, if the UE support 4 ports, gNB will us use 4 ports to communicate with the UE.
Yes, and most beamforming deployments use TDD.
Perhaps in future if there are business need we may see mmWave based indoor solutions like Ericsson has 3.5 GHz capable Radio Dots.
Could be, but in indoor WIFI is king.
I do not know why any business would pay over 8-10K EURO to deploy a small cell like radio dots of Ericsson.
Even LAN cable is king in indoor.
I believe that Japanese MNOs are deploying mmWave on a relatively large scale.
This information is as of February 2023.
mmWave UEs have a small MaMIMO array. Typically 2-3 such arrays which are on different sides of the UE to prevent “devastating” blockage you get when you put your finger/hand over it.
Check for example iPhone patents, they show locations of those beamforming arrays.
Why only 2 layers are supported at mmWave UE? That’s because those arrays are using analogue beam forming therefore you can have only one layer per polarisation.
1 Like