3A Interoperability with different IP-CAN

24.2293GPPIP multimedia call control protocol based on Session Initiation Protocol (SIP) and Session Description Protocol (SDP)Release 18Stage 3TS

The IM CN subsystem can be accessed by UEs resident in different types of IP-CAN. The main body of this document, and annex A, are general to UEs and IM CN subsystems that are accessed using any type of IP-CAN. Requirements that are dependent on the type of IP-CAN are covered in annexes B, E, H, L, M, O, Q, R, S, U and W.

At any given time, for a given SIP transaction or dialog, the UE sees only one type of IP-CAN, as reported to it by the lower layers. The UE follows the procedures of the IP-CAN specific annex related to the last type of IP-CAN reported, even if it is different to one used previously. In particular, handover at the radio layers between two different access technologies can result in such a change while the dialog or transaction proceeds.

At any given time, for a given SIP transaction or dialog, the P-CSCF sees only one type of IP-CAN, as determined by interface to a particular resource architecture, e.g. policy and charging control, and by the access technology reported to it over that interface, or in the absence of this, by preconfiguration in the system. The P-CSCF follows the procedures of the IP-CAN specific annex related to the last type of IP-CAN determined, even if it is different to one used previously. In particular, handover at the radio layers between two different access technologies can result in such a change while the dialog or transaction proceeds.

It is the responsibility of the IP-CAN to ensure that usage of different bearer resources are synchronised on the handover from one IP-CAN to another, e.g. so that a signalling bearer provided by one IP-CAN is a signalling bearer (if provided by that IP-CAN) after handover, and that the appropriate QoS and resource reservation exists after handover. There is no SIP signalling associated with handover at the IP-CAN, and therefore no change in SIP state at one entity is signalled to the peer SIP entity when handover occurs.

In particular the following constraints exist that can have an impact on P-CSCF usage:

1) some IP-CANs can explicitly label a bearer as a signalling bearer, while others provide a bearer that has appropriate QoS, but no explicit labelling. Therefore if handover occurs from an IP-CAN with explicit labelling, to an IP-CAN with no explicit labelling, and then back to an IP-CAN with explicit labelling, the signalling will then be on a bearer that is not explicitly labelled; and

2) some IP-CANs support signalling of grouping of media within particular bearers, while others do not. Therefore if handover occurs from an IP-CAN with grouping, to an IP-CAN with no grouping, and then back to an IP-CAN with grouping, the signalled grouping can have been lost.

When a UE supports multiple IP-CANs, but does not support handover between those IP-CANs, the annex specific to that IP-CAN applies unmodified.

Where handover between IP-CANs occurs without a reregistration in the IM CN subsystem, the same identies and security credentials for access to the IM CN subsystem are used before and after the handover.

At the P-CSCF, the access technology can variously use the PCRF or PCF or NASS in support of both signalling and media bearer provision (or indeed use neither). How to determine which applies is up to network dependent rules, but can be specific to the access technology used by each different UE. Not all access technologies are defined for use with NASS, and not all access technologies are defined for use with the PCRF or PCF.