Dear all, I reviewed draft-clt-dmm-tn-aware-mobility-02 and I have the following comments based on 3GPP architectural definitions in TS 23.501 / 23.502
1. Section 2.1 - RQI is just a flag to tell the UE that it has to reflect back the same QFI in the uplink packet. RQI on its own is not a QoS indicator. Hence the use of RQI in the following sentence is not correct Mapping of the PDU sessions to TE paths can be done based on the source UDP port ranges (if these are assigned based on the PDU session QCIs, as done in some deployments with 4G/LT) of the GTP-U encapsulated packet or based on the 5QI or RQI values in the GTP-U header. 2. Section 2.1, 2.2 - and almost everywhere - Restricting to description of gNB as the radio side is not correct. 5G architecture allows the radio or access node to be any of the following: a) gNB (which means radio is NR) b) Ng-eNB (which means radio is EUTRA) c) Untrusted WLAN d) From R16 onwards wireline and Trusted wireless LAN access So the draft should generally mention as 5G-AN and not as gNB. 3. The draft mixes SSC modes with slicing concept. SSC modes and slicing are two independent concepts. UL/CL and BP UPF are applicable to all SSC modes. SSC modes do not have any relevance when selecting a transport path. The description of SSC modes need to be corrected to rather reflect how / where the PPR-ID is applied for different SSC modes. Regards Sridhar Bhaskaran _______________________________________________ dmm mailing list dmm@ietf.org https://www.ietf.org/mailman/listinfo/dmm