Content for TS 23.436 Word version: 19.3.0

0… 5… 6… 8… 8.3… 8.4… 8.5… 8.6… 8.7… 8.8… 8.9… 8.10… 8.11… 8.12… 8.13… 8.14… 8.15… 8.16… 8.17… 8.18… 8.19… 9… 9.3… 9.4… 10… A…

8.18 Procedure for supporting DN Energy Efficiency analytics 8.18.1 General 8.18.2 Procedure 8.18.3 Information flows
...

8.18 Procedure for supporting DN Energy Efficiency analytics |R19| p. 107

8.18.1 General p. 107

This clause describes the procedure for DN energy consumption/efficiency analytics, where the analytics are performed based on data collected from one or more DNs and A-ADRF.

8.18.2 Procedure p. 107

Figure 8.18.2-1 illustrates the procedure for DN energy efficiency analytics enablement solution.

Pre-conditions:

Data producers (e.g. A-ADRF, EAS, EES) may be pre-configured with data producer profiles (as in Table 8.2.4.8-1) for the data they can provide. ADAES has discovered available data producers and their data producer profiles.

Reproduction of 3GPP TS 23.436, Fig. 8.18.2-1: ADAES support for DN energy analytics

Figure 8.18.2-1: ADAES support for DN energy analytics

Step 1.

The VAL server sends a DN energy analytics request to ADAES to perform analytics on the DN Energy Consumption/Efficiency for one or more DNs/EDNs, Event ID= "DN energy analytics", for a given DN service area (or subarea) and a given time window.

Step 2.

The ADAES authorizes the VAL request.

Step 3.

The ADAES requests and receives from the EAS /VAL servers hosted at the serving and target DNs (within the VAL service area), expected application service load and traffic schedules for the ongoing or future sessions within the area. Such data include traffic schedule report for the VAL Server, and this step re-uses the step 3 to 10 of clause 8.8.2.1.

Step 4.

The ADAES calculates the expected energy consumption or efficiency based on the received traffic and load data for the given DNN/DNAI based on the request.

Step 5.

The ADAES obtains the corresponding trained ML model based on procedure in clause 8.3.2 of TS 23.482 and performs analytics to derive the predicted energy consumption at the target area and time horizon. The analytics outputs can be the predicted energy consumption / efficiency for the given DNN/DNAI.

Step 6.

The ADAES sends a DN energy analytics response with the energy consumption/efficiency analytics output data to the VAL server.

Based on 6, the VAL server can use these analytics as input to trigger pro-actively:

an application server migration to a different edge cloud or to a centralized cloud as a way of reducing the energy consumption for the edge (if consumption is expected to be very high (e.g. higher than a pre-configured threshold)).
an application server offboarding and the instantiation of a new server at the target edge/centralized cloud to minimize energy consumption of the edge platform (taking into account the system wide energy efficiency).

8.18.3 Information flows p. 109

8.18.3.1 General p. 109

The following information flows are specified for DN energy analytics based on clause 8.18.2.

8.18.3.2 DN energy analytics request p. 109

Table 8.18.3.2-1 describes information elements for the DN energy analytics request from the VAL server to the ADAE server.

Table 8.18.3.2-1: DN energy analytics request

Information element	Status	Description
Analytics Consumer ID	M	The identifier of the analytics consumer (VAL server, EAS).
Analytics ID	M	The identifier of the analytics event. This ID can be for example "edge performance analytics".
Analytics type	M	The type of analytics for the event, e.g. statistics or predictions.
DNN/DNAI	M	DNN or DNAIs information for which the subscription applies.
Energy Efficiency/Consumption metrics	O	The formula and necessary metrics for calculating the EE based on load and traffic information per DN.
Target data producer profile criteria	O	Characteristics of the data producers to be used.
Preferred confidence level	O	The level of accuracy for the analytics service (in case of prediction).
Area of Interest	O	The geographical or service area for which the subscription request applies.
Time validity	O	The time validity of the subscription request.
Reporting requirements	O	It describes the requirements for the energy analytics reporting. This requirement may include e.g. the type and frequency of reporting (periodic or event triggered), the reporting periodicity in case of periodic, and reporting thresholds.

8.18.3.3 DN energy analytics response p. 109

Table 8.18.3.3-1 describes information elements for the DN energy analytics request from the VAL server to the ADAE server.

Table 8.18.3.3-1: DN energy analytics response

Information element	Status	Description
Result	M	The result of the analytics request (positive or negative acknowledgement).
Analytics ID	O	The identifier of the analytics event.
Analytics Output	O	The predictive or statistical parameter, which can be stats or prediction related to the energy consumption or efficiency for the edge platform for a given area/time and based on the event type.
> DNN	M	Identifies the data network name for which analytics information is provided.
> DNAI	M	Identifier of a user plane access to one or more DN(s) of the DN.
> Energy metrics	O	The predicted energy metrics.
>> Energy Consumption (NOTE)	O	The predicted energy consumption per DNAI based on network and edge resource usage.
>> Energy Efficiency (NOTE)	O	The energy efficiency per DNAI based on network and edge resource usage (given a certain optimal energy consumption metric, which can be pre-configured).
>> DN Data Volume (NOTE)	O	The predicted data volume per DNAI.
> Area of Interest	O	The area (topological or geographical or edge area) where the analytics apply.
> Applicable time period	O	The time period that the analytics applies to.
> Confidence level	O	For predictive analytics, the achieved confidence level can be provided.
NOTE: At least one of these shall be present.