“If the MAC entity has UL resources allocated for a new transmission the MAC entity shall, . . . , if the Power Headroom reporting procedure determines that at least one PHR has been triggered and not cancelled; and if the allocated UL resources can accommodate the MAC CE for PHR which the MAC entity is configured to transmit, plus its subheader, as a result of LCP as defined in subclause 5.4.3.1, . . . , instruct the Multiplexing and Assembly procedure to generate and transmit the Multiple/Single Entry PHR MAC CE as defined in subclause 6.1.3.9/6.1.3.8 based on the values reported by the physical layer”;
it is not really/precisely specified in which PUSCH the PHR MAC CE is included. This specification text is intentionally formulated a bit imprecise so that it leaves some freedom for the UE implementation where to transmit the PHR MAC CE, such as for example, which PUSCH resource to use for PHR MAC CE inclusion. This has been also confirmed in a 3GPP RAN2 meeting (common understanding among companies in RAN2). For the scheduling performance, it is not very important on what exact PUSCH resource the PHR MAC CE is transmitted as long as it is clear to the eNB/gNB what information (e.g. DCI, CG PUSCH allocations, TPC commands) UE used for the calculation of the PHR. The network needs to know on what basis the PHR calculation was done.