The steps involved in the handover from 3GPP Access connected to the EPC to trusted non-3GPP IP access are depicted below for the case of non-roaming, roaming with home routed traffic, roaming with local breakout and roaming with anchoring in the Serving Gateway in the VPLMN. It is assumed that while the UE is served by the 3GPP Access, a PMIPv6 or GTP tunnel is established between the S-GW and the PDN-GW in the evolved packet core.
This procedure supports the home routed (Figure 4.2.2-1), roaming (Figure 4.2.3-1) and Local breakout (Figure 4.2.3-4) case. The PCRF in the HPLMN is informed of the change and any change in the policy that results is signalled to the Serving-GW. The signalling takes place through the vPCRF in the VPLMN. In the case of Local Breakout, the PDN-GW in the VPLMN exchanges messages with the vPCRF.
The optional interaction steps between the gateways and the PCRF in Figure 8.2.2-1 only occur if dynamic policy provisioning is deployed. Otherwise policy may be statically configured with the gateway.
For connectivity to multiple PDNs the following applies:
If the UE is connected to both 3GPP access and non-3GPP access before the handover of PDN connections to trusted non-3GPP access is triggered, steps 2 to 10 shall be skipped and the UE shall only perform step 11 for each PDN connection that is being transferred from 3GPP access.
If the UE is connected only to 3GPP access before the handover of PDN connections to trusted non-3GPP access is triggered, steps 2 to 10 shall be performed. In step 4 the UE shall provide an APN corresponding to one of the PDN connections that are being transferred from 3GPP access. The UE shall then repeat step 11 for each of the remaining PDN connections that are being transferred from 3GPP access.
Step 12 shall be repeated for each PDN connection that is being transferred from 3GPP access.
Step 11 can occur in parallel for each PDN. Other impacts related to the handover for multiple PDNs are described in clause 8.1.
The UE discovers the trusted non-3GPP IP access system and determines to transfer its current sessions (i.e. handover) from the currently used 3GPP Access to the discovered trusted non-3GPP IP access system. The mechanisms that aid the UE to discover the trusted non-3GPP IP access system, are specified in clause 4.8 (Network Discovery and Selection).
The UE performs access authentication and authorization in the non-3GPP access system. The 3GPP AAA server authenticates and authorizes the UE for access in the trusted non-3GPP system. The 3GPP AAA server queries the HSS and returns the PDN-GW identity or identities to the trusted non-3GPP access system at this step (upon successful authentication and authorization). The 3GPP AAA Server also returns to the trusted non-3GPP access system the MN NAI to be used to identify the UE in Proxy Binding Update and Gateway Control Session Establishment messages (steps 5 and 6).
PDN-GW address selection is as described in clause 4.5.1 of this specification. The PDNs the UE is connected to before handover are obtained from the HSS with the UE subscriber data.
After successful authentication and authorization, the L3 attach procedure is triggered. At the latest, in this step, the UE should indicate its capability for the IP address preservation. How this information is signalled from the UE to the access network is outside of the scope of 3GPP.
If the UE provides an APN, the Trusted non-3GPP Access verifies that it is allowed by subscription. If the UE does not provide an APN, and the subscription context from HSS contains a PDN-GW identity and APN pair corresponding to the default APN, the Trusted non-3GPP Access uses the default APN. The case where the APN selected for the handover attach (default APN or the APN provided by the UE) does not have corresponding PDN-GW identity information in the subscription context is considered as an error case.
The Trusted Non-3GPP IP Access initiates a Gateway Control Session Establishment Procedure with the PCRF as specified in TS 23.203. If the Trusted Non-3GPP IP Access supports UE/NW bearer control mode, the PCRF provides all the QoS rules required for the Trusted Non-3GPP IP Access to perform the bearer binding.
If the updated rules require network-initiated dynamic resource allocation for the UE, the resource allocation takes place before step 6.
If the Handover Indicator in the Proxy Binding Update (to be sent in step 6) is set to indicate either initial attach or that the handover state is unknown, the Trusted non-3GPP IP Access indicates in the Gateway Control Session Establishment message that linking with the Gx session shall be deferred until step 7, as specified in TS 23.203. In this case, when performing the leg linking, the PCRF verifies that the IP-CAN type reported over Gxa and Gx are the same.
The entity in the Trusted non-3GPP IP Access acting as a MAG sends a Proxy Binding Update (MN-NAI, Lifetime, Access Technology Type, Handover Indicator, APN, GRE key for downlink traffic) message to the PDN-GW in order to establish the new registration. The MN NAI identifies the UE for whom the message is being sent. The Lifetime field must be set to a nonzero value in the case of a registration. Access Technology Type is set to a value matching the characteristics of the non-3GPP access. The APN may be necessary to differentiate the intended PDN from the other PDNs supported by the same PDN-GW. The MAG creates and includes a PDN connection identity if the MAG supports multiple PDN connections to a single APN.
The PDN-GW executes a PCEF-Initiated IP CAN Session Modification Procedure with the PCRF as specified in TS 23.203. The Event Report indicates the change in Access Type.
If the PDN-GW decided to allocate a new IP address/prefix instead of preserving the old IP address/prefix, as described in clause 18.104.22.168.3, the PDN-GW executes an IP-CAN session Establishment Procedure with the PCRF instead of a PCEF-Initiated IP-CAN Session Modification Procedure.
The PDN-GW informs the 3GPP AAA Server of its PDN-GW identity and the APN corresponding to the UE's PDN Connection and obtains authorization information from the 3GPP AAA Server. The message includes information that identifies the PLMN in which the PDN-GW is located. The 3GPP AAA Server may update the information registered in the HSS as described in clause 12.
The PDN-GW responds with a PMIP Binding Acknowledgement (MN NAI, Lifetime, UE Address Info, Additional Parameters, GRE key for uplink traffic, Charging ID) message to the Trusted Non-3GPP IP Access. The MN NAI is identical to the MN NAI sent in the Proxy Binding Update. The Lifetime indicates the duration the binding will remain valid. If the corresponding Proxy Binding Update contains a PDN connection identity, the PDN-GW shall acknowledge if the PDN-GW supports multiple PDN connections to a single APN. The UE address info returns the IP Address assigned to the UE. The optional Additional Parameter information element may contain other information. Since this step is triggered by the Proxy Binding Update message from the Trusted non-3GPP IP Access in step 6 and the result of the optional step 7, it can occur after step 7. If step 7 is not taken, this step can occur after step 6. The Charging Id provided is the Charging Id previously assigned to the PDN connection if the source access is a PMIP-based access or to the Default Bearer if the source access is GTP-based.
For connectivity to multiple PDNs, the UE establishes connectivity to all the PDNs that are being transferred from 3GPP access besides the PDN connection that was established in the steps 3-10, as described in clause 6.8.1.
The PDN-GW shall initiate the PDN-GW Initiated PDN Disconnection procedure in 3GPP access as defined in clause 22.214.171.124 or the PDN-GW Initiated Bearer Deactivation procedure as defined in TS 23.401, clause 126.96.36.199.