Optimization assistance information for the Radio Access Network (RAN) in 5GC

I. Core Network Assistance Information in 5G: This is designed to assist RAN in optimizing User Equipment (UE) state transition control and RAN paging strategies in the RRC Inactive state. Core network assistance information includes the information set "Core Network Assisted RAN Parameter Tuning," which helps the RAN optimize UE RRC state transitions and CM state transition decisions. It also includes the information set "Core Network Assisted RAN Paging Information," which helps the RAN develop optimized paging strategies when RAN paging is triggered.

II. Core Network Assisted RAN Parameter Tuning helps the RAN minimize UE state transitions and achieve optimal network behavior. The current specifications do not define how the RAN uses core network assistance information.

• Core network assisted RAN parameter tuning can be tuned by the AMF for each UE based on collected UE behavior statistics, expected UE behavior, and/or other available information about the UE (e.g., subscribed DNN, SUPI range, or other information).
• If the AMF maintains expected UE behavior parameters, network configuration parameters (as described in TS 23.502 [3] sections 4.15.6.3 or 4.15.6.3a), or SMF-derived core network assisted RAN parameter tuning, the AMF can use this information to select core network assisted RAN parameter values. If the AMF can derive the UE's mobility pattern (as described in section 5.3.4.2), the AMF can consider mobility pattern information when selecting core network assisted RAN parameter values.
• The SMF uses SMF-associated parameters (e.g., UE's expected behavior parameters or network configuration parameters) to derive SMF-derived CN assisted RAN parameter tuning. The SMF sends the SMF-derived CN assisted RAN parameter tuning to the AMF during the PDU session establishment process. If the SMF-associated parameters change, the PDU session modification procedure is applied. The AMF stores the SMF-derived CN assisted RAN parameter tuning in the PDU session level context. The AMF uses the SMF-derived CN assisted RAN parameter tuning to determine the PDU session level "expected UE activity behavior" parameter set, which may be associated with the DU session ID, as described below.
• Expected UE behavior parameters or network configuration parameters can be provided to the AMF or SMF by an external party via the NEF, as described in Section 5.20.

III. CN-assisted RAN parameter tuning provides the RAN with methods to understand UE behavior, specifically including the following aspects:

• "Expected UE activity behavior," which refers to the expected pattern of UE switching between CM-CONNECTED and CM-IDLE states, or the duration of the CM-CONNECTED state. This can be obtained from sources such as statistical information, expected UE behavior, or user information. The AMF derives one or more sets of "expected UE activity behavior" parameters for the UE as follows:
• The AMF can derive and provide the RAN with a UE-level set of "expected UE activity behavior" parameters, which considers the expected UE behavior parameters or network configuration parameters received from the UDM (see Sections 4.15.6.3 or 4.15.6.3a of TS 23.502 [3]) and the SMF for CN-assisted RAN parameter tuning. Control plane CIoT 5GS optimization is used to tune parameters related to PDU sessions. This set of "expected UE activity behavior" parameters is valid for the UE; and
• The AMF can provide the RAN with a PDU session-level set of "expected UE activity behavior" parameters, for example, considering CN-assisted RAN parameter tuning derived from the SMF, for each established PDU session.

IV. The PDU session-level "expected UE activity behavior" parameter set is associated with and valid for the PDU session ID. The RAN can consider the PDU session-level "expected UE activity behavior" parameters when the user plane resources of the PDU session are activated;

• "Expected handover behavior," which refers to the expected interval between inter-RAN handovers. This can be derived by the AMF, for example, from mobility pattern information;
• "Expected UE mobility," which indicates whether the UE is expected to be stationary or mobile. For example, this information can be obtained from the following sources: statistical information, expected UE behavior parameters, or subscription information;
• Expected UE mobility trajectory, for example, can be obtained from statistical information, expected UE behavior parameters, or subscription information; or
• UE differentiation information includes expected UE behavior parameters, but does not include the expected UE mobility trajectory (see clause 4.15.6.3 of TS 23.502 [3]), to support Uu operation optimization for NB-IoT UE differentiation (if the RAT type is NB-IoT).

----The AMF decides when to send this information as "expected UE activity behavior" to the RAN via an N2 request, through the N2 interface (see TS 38.413 [34]).

----The CN assisted information calculation, i.e., the algorithm used and relevant criteria, and the decision on when it is considered appropriate and stable to send it to the RAN, are vendor-specific.