How is spatial diversity simulated in 5G-LENA?

[Alan Birchler]

Hello everyone,

From my basic understanding of wireless communication theory, spatial multiplexing is related to splitting an information stream along multiple paths while spatial diversity is related to copying a stream across different paths. If these assumptions are correct, how does 5G-LENA simulate spatial diversity? 

From what I gathered in the documentation, the MIMO classes are related to spatial multiplexing to improve spectral efficiency and throughput. Does MIMO, in the way it's implemented in 5G-LENA, also handle spatial diversity or is this handled in another component? If the latter is correct, if in my simulation I don't enable MIMO (which I'm assuming means setting NrHelper::EnableMimoFeedback to false) but I do configure multiple antennas on both UEs and gNodeBs, is spatial diversity still performed/simulated? Is there anything else I need to enable to simulate this aspect? 

Best regards,

Alan Birchler De Allende

[Sandra Lagén]

Hi Alan,

Multiple tx/rx antennas can be exploited in different ways, as you said: 1) spatial multiplexing (multiple streams over the same channel), and 2) spatial diversity (same stream over multiple paths). We support both, but for the later (spatial diversity) just in the form of transmit/receive beamforming. For 1), you have to enable MIMO feedback and also you need to configure the tx/rx antenna arrays with multiple ports. The number of streams that can be sent is limited by the minimum between tx/rx antenna ports. You can enable multiple ports either by configuring dual polarization or by configuring horizontal/vertical ports in the antenna array of tx&rx. If, on the other hand, for 2), you don't enable MIMO feedback, and you don't want to have SU-MIMO, you just need 1 port, and then the multiple antenna elements will be used to concentrate the tx power, as well as the rx combining gain, into the desired direction, through tx/rx beamforming. 

Hope it helps!

BR,

Sandra

Missatge de Alan Birchler <albir...@gmail.com> del dia dj., 23 d’oct. 2025 a les 12:59:

--
You received this message because you are subscribed to the Google Groups "5G-LENA-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 5g-lena-user...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/5g-lena-users/67807544-d676-4ac9-846d-8f6f64512a1an%40googlegroups.com.

[Alan Birchler]

Hi Sandra,

Thanks for the prompt rely and the clear explanation!

In regards to beamforming, if I'm using an instance of IdealBeamformingHelper as the Beamforming model within my simulation, does this imply that communication between a gNodeB and a UE within my simulation will communicate at the most optimal spatial diversity within the set values of other simulation parameters (number of antennas, distance, etc.)?

Best regards,

Alan Birchler De Allende

[Sandra Lagén]

The "most optimal" depends how you define optimal... Within the ideal BF, the actual method is set by the BeamformingMethod attribute. You can find more details in the documentation: https://cttc-lena.gitlab.io/nr/nrmodule.pdf (Section 2.3.9). CellScan selects the best beam within a subset of predefined beams, while DirectPath points the beam into the LOS path of the intended receiver. So, we don't have the optimal BF design for a given channel, but those implemented approximate actual practical procedures, based on a codebook of beams or based on AoA/AoD estimates.

BR,

Sandra

Missatge de Alan Birchler <albir...@gmail.com> del dia dj., 23 d’oct. 2025 a les 14:33:

To view this discussion, visit https://groups.google.com/d/msgid/5g-lena-users/53934ba6-f2d5-48e6-ace6-643ac73401fbn%40googlegroups.com.