Content for TS 28.310 Word version: 18.5.0

1… 4… 5… 6… 6.2… A…

4.1 EE KPIs Overview 4.2 Management services 4.3 Energy saving
...

4 Concepts and overview p. 10

4.1 EE KPIs Overview p. 10

Telecommunication networks energy efficiency KPIs are defined by various SDOs / organizations and are of various natures. They can be applied to either:

whole networks (i.e. end-to-end), or to
sub-networks (e.g. the radio access network), or to
single network elements, or to
telecommunication sites, which contain network elements and site equipment.

Moreover, EE KPIs can also be categorized according to the operator's network life cycle phase they may apply to, e.g.:

during the Buy phase, mobile network operators may be willing to compare network elements from various vendors from an EE standpoint. Some EE KPIs and measurement methods have been specified for this purpose.
during the Design / Build phase, mobile network operators are always faced to several design options, and may be willing to compare them from an EE standpoint. This may happen for the whole network, sub-networks and for telecom sites. For telecom sites, EE KPIs have been specified.
during the Run phase, mobile network operators need to assess the energy efficiency of the live network, as a whole (i.e. end-to-end), or for sub-networks, or for single network elements or telecom sites. Some EE KPIs and measurement methods have also been specified for this purpose.

Generally, EE KPIs for network elements are expressed in terms of Data Volume divided by the Energy Consumption of the considered network elements. In the case of radio access networks, an EE KPI variant may also be used, expressed by the Coverage Area divided by the Energy Consumption of the considered network elements.

The calculation of the energy efficiency of 5G networks relies on the following principles:

it is based on the two high-level EE KPIs defined in ETSI ES 203 228 [2]:
- EE_MN,DV = DV_MN / EC_MN, and
- EE_MN,CoA = coverage area / EC_MN
EE_MN,DV may apply to the whole 5G network whereas EE_MN,CoA may apply only to NG-RAN;
EE_MN,DV requires the collection of both Data Volumes (DV) and Energy Consumption (EC) of 5G Network Functions (NF);
In NG-RAN, DV is measured per cell;
In 5GC, DV is measured per NF;
EC definition and measurement method for 5G PNFs rely on ETSI ES 202 336-1 [3] and ETSI ES 202 336-12 [4];
EC is measured by PEE parameters (cf. ETSI ES 202 336-12 [4] - Annexes A and B);
PEE measurements requirements for all deployment scenario in NG-RAN: The 3GPP management system responsible for the management of the gNB (single or multiple vendor gNB) shall be able to collect PEE measurements data from all PNFs in the gNB, in the same way as the other PM measurements;
When gNBCU/gNBCU-CP/gNBCU-UP energy consumption is assumed to be very small compared to gNBDU and given that, in some cases, the gNBCU/gNBCU-CP/gNBCU-UP may be virtualized, the present document only considers the energy consumed in gNBDU(s) (in case of split scenarios) and in non-split gNBs (see clause 4.2.1 of TS 28.541 and clause 6.1.1 of TS 38.401). There might be a need for some correction in KPI between the different deployment scenarios.

NOTE 3:
The vendor(s) of 2-split (gNBDU/gNBCU) or 3-split gNB/en-gNB components (gNBDU/gNBCU-CP/gNBCU-UP) may be same or different depending on the implementations.
In the present document, it is assumed that NG-RAN is only composed of base stations with built-in sensors (see ETSI ES 202 336-12 [4] - clause 4.4.1).

Besides the parameters required to calculate the energy efficiency, e.g. DV and EC, other parameters may be used to interpret variations in energy efficiency KPI values from different networks. These parameters can be classified into demography, topography and climate classes (see ETSI ES 203 228 [2] - section 4.3), which describe the network characteristics with regard to population density, geographical conditions and climate zones. For each class of parameters, there can be subclasses, e.g. demography can be further classified into dense urban, urban, sub-urban, rural or unpopulated scenarios. For each class / subclass, the energy efficiency KPI values may be interpreted differently.

4.2 Management services p. 11

The management services required for the assessment of the energy efficiency of 5G networks are listed below:

Performance management services (see clause 4.3 of TS 28.550):
- Measurement job control service for NF.
- Performance data file reporting service for NF.
- Performance data streaming service for NF.
Management services for network function provisioning (cf. clause 6.3 of TS 28.531):
- Provisioning for NF.
- Provisioning data report for NF.
Management services for Fault Supervision (cf. clause 4.1.1 of TS 28.545):
- Fault supervision data report service for NF.
- Fault supervision data control service for NF.

4.3 Energy saving p. 12

4.3.1 Introduction p. 12

Operators are aiming at decreasing power consumption in 5G networks to lower their operational expense with energy saving management solutions. With the foreseen deployment of more NR base stations, e.g., small base stations with massive MIMO in high-band, energy saving becomes even more urgent and challenging.

Management of 5G networks contributes to energy saving by reducing energy consumption of 5G networks, while maintaining coverage, capacity and quality of service. The permitted impact on coverage, capacity and quality of service is determined by operator's decision.

4.3.2 Concepts p. 12

Two energy saving states can be conceptually identified for cells, NEs and NFs.

Conceptually, a cell or a network element or network function may be on one of these two states with respect to energy saving:

notEnergySaving state
energySaving state

Based on the above energy saving states, a full energy saving solution includes two elementary procedures:

Energy saving activation (change from notEnergySaving state to energySaving state)
Energy saving deactivation (change from energySaving state to notEnergySaving state)

When a cell is in energy saving state it may need candidate cells to pick up the load. However a cell in energySaving state should not cause coverage holes or create undue load on the surrounding cells. All traffic on that cell is expected to be drained to other overlaid/umbrella candidate cells before the cell moves to energySaving state.

Similarly, when a network element or network function is in energy saving state it may need candidate network elements or network functions to pick up the load. For example, during off-peak traffic periods, one or more edge UPFs in energySaving state should not cause undue load on the other UPFs, all remaining traffic on those edge UPFs is expected to be redirected to other UPFs before the edge UPFs move into energySaving state.

A cell in energySaving state is not considered as a cell outage or a fault condition. No alarms should be raised for any condition that is a consequence of a subject cell or network element or network function moving into energySaving state.