RSRP Measurement Granularity in Sidelink Mode 2 Implementation

[Andres Calle]

According to the 3GPP specification for resource allocation in sidelink mode 2, the UE is expected to measure the RSRP per subchannel in each time slot, as this information is essential for efficient resource selection based on perceived interference. However, when reviewing the implementation in 5G-LENA, I noticed that in the nr-ue-phy.cc file, only a single RSRP value is calculated per slot, without distinguishing between subchannels. Furthermore, in nr-sl-ue-mac.cc, it also seems that only one RSRP value per slot is used for resource selection.

Does this implementation deviate from the behavior specified in the standard? Or am I missing something in the logic of the mode 2 implementation that justifies the use of a single RSRP value per slot?

Any clarification on how this aspect is handled in 5G-LENA and whether there are plans to align it more closely with the 3GPP standard would be greatly appreciated.

[Gabriel Ferreira]

I haven't touched the sidelink codebase, so I can't answer on that.
On the the upstream code, do not use the RSRP. It has not been validated, nor is according to 5G-NR standard.
We do not claim to support it just yet, same for handover. We are on that.
This was upstreamed because splitting the code and porting upper layers step by step would be more work.

[Gabriel Ferreira]

*We are working on that.