5GC as Network as a Service (NaaS) can enable an Operator to connect his existing or 5G radio to delivery telecommunication service to end user. Network Function Services, on the other hand, are just service flows supported by different network elements that individually have no relevance. 3GPP definition of service seems heavily loaded by two aspects - interactions with other network functions and underlying protocol. The reference point representation is most explicit mapping of 5G to LTE product based architecture.
Furthermore, most vendors see this representation as evolution of legacy product portfolio for network functions realization. Most incumbent vendors have already started early discussions with Operators making one think of 5G as product evolution than new service introduction. This has triggered Operator RFI/RFPs to include product dimensioning inputs as mandatory sections as part of answering. However, there's one problem that seem to be forgotten - These network functions are merely software deliverables and have no hardware associated unlike there legacy parts. With no standardized resource size, there's no way to compare performance of network function coming from multiple vendors.
Also, each Operator will have their own common cloud environment for all NFs. Operator will thus be responsible for underlying data centers, cloud management and orchestration (using Kubernetes) and NF OAMs. Vendor role is limited to ensuring support for cloud environment and delivery of software bundle for network function. Resource sizing is Operator prerogative. Which bring us the actual question that needs answering -
Is dimensioning of 5G network function useful and should Operator really be asking for it ?
We think Operators should instead focus on standard or recommended resource sizing inputs from vendors for maximizing SW performance. Calculating transaction capacity for each interface is pointless exercise with orchestration supporting dynamic scaling of network functions based on utilization or other KPI inputs. As we progress in our 5G journey and more operators will realize futility of this exercise. What bothers us is the missed opportunity by Operators to revisit #5Glicensing because of legacy concepts like this one.
We would love to hear your views. Also, check out our related blog on 5G Product Licensing.