Content for  TR 22.882  Word version:  19.3.0

One of the strategies to save energy within mobile networks is to shut down some RAN nodes at times of low usage. Eventually only one communication service would be used. Thus, there is a potential for further gain to be exploited by pooling the communication service on a local basis among operators at times of low usage. Agreements could be put in place between operators so that in the low load periods (e.g., night time) only one of multiple mobile networks may be active in an area and will provide communication service to the subscribers of all networks, whereas the other networks can apply cell shutdown of their own infrastructure to obtain network energy savings. Alternatively, based on risks of power outage nationwide/regionwide, regulators could ask operators to "optimize" their coverage e.g., shutdown some nodes in overlapping coverage areas during energy peak hours and/or in specific geographical areas, whilst still guaranteeing minimum coverage/service (in particular emergency calls). This can also apply between NPN operators and/or with PLMN operators.

• OP1 and OP2 are two PLMN operators.
• There is an overlap coverage between OP1 and OP2, which both provide mobile communication services to their subscribers on various bands.
• There are mutual agreements between OP1 and OP2 allowing them to provide communication services to the subscribers of the other network, in case it is not active in an area for low load. They define e.g. on a daily basis or specific locations a time when the communication service pooling starts and ends, and can include other parameters like preferred bands etc.
• OP3 operates an NPN dedicated to a factory around its campus, which is mainly used for IIoT purposes, but also for employees.
• There is a business agreement between the OP1 and OP3, i.e. OP3 users can be served by OP1 network (but not the other way around) based on certain conditions. At night, OP3 shuts down its network when the machines are off, as the little remaining traffic is generated by some employees staying late or overnight. The agreement requests OP1 to provide access to OP3 UEs during the night hours for this type of traffic.
• UE 1 belongs to OP1. UE 2 belongs to OP2. UE 3 belongs to OP3.

1. At 8PM, OP3 ("beneficiary" network) starts informing its currently served UEs within a specific area that it will shut down its network and request them to move to OP1 ("donor" network). It can also indicate the time when it will resume connectivity (e.g. 8AM).
2. OP1 accepts OP3's UEs onto its network based on "EE-based communication service pooling" reason (it wouldn't have without agreement).
3. Once OP3 detects no UE is served anymore on its network, it shuts down its network
4. On the next morning at 8AM, OP3 powers up its network again in that area
5. OP3 UEs return to OP3
6. OP1 stop serving OP3 UEs.

Furthermore, as this is an industrial campus area, traffic for OP2 is low this Saturday (only OP3 factory is working). Based on its EE KPIs in that area and according to the agreement with OP1, OP2 decides to shut down its cells until Monday morning 6AM with the same mechanism. OP1 starts serving OP2 UEs under its own network during this time. In this case the decision is dynamic and not only based on fixed times, but on other conditions, within the agreed conditions between operators (e.g., anytime during weekends).

After OP2 and OP3 have shut down their networks, their subscribers can still be served via OP1. OP2 subscribers under "EE-based communication service pooling" are not charged differently when served by OP1 network, with respect to when they are under OP2 network coverage. OP2 may be charged by OP1 as per their company agreement, e.g. based on a flat cost, per subscriber, data volume, duration etc. OP3 may also be charged by OP1 as per their company agreement.

Network sharing is an existing technique used to save resources across operators (see clause 6.21 of TS 22.261), which could be leveraged for "communication service pooling" for energy saving purposes. However, current network sharing agreements are mainly on a permanent basis with little flexibility in time and space. Indirect network sharing is a promising technique that can be considered for this use case. Minimization of Service Interruption (MINT) as defined in clause 6.31 of TS 22.261, in another existing feature, which has specified that "UEs can obtain service in the event of a disaster, if there are PLMN operators prepared to offer service. The minimization of service interruption is constrained to a particular time and place. To reduce the impact to the 5G System and EPS of supporting Disaster Roaming, the potential congestion resulting from an influx or outflux of Disaster Inbound Roamers is taken into account.". Requirements exist, e.g., "to provide means to enable a UE to access PLMNs in a forbidden PLMN list if a Disaster condition applies and no other PLMN is available except for PLMNs in the forbidden PLMN list". Disaster roaming is further specified in clauses 4.4.3.3.1 and 3.10 of TS 23.122. However, this use case is not related to disaster condition. Furthermore, differently from disaster roaming, there may be no detection of failure of home PLMN by the UE, and the pooling (i.e., roaming) duration may be known in advance.

[PR.5.11.6-1] [PR.5.11.6-2] [PR.5.11.6-3] [PR.5.11.6-4]