What should be the main goal of any 5G architecture considering the network, service, and business, as well as operation, control and management.
Is it to realise a largely self-governed network that adjusts its behaviour according to service requirements, business requirements, and instantaneous demand?
I think those goals provide the frame for how the network must be configured (operator goals), what it should provide (user goals), how it provides that (service goals) and how it can be managed (control and management goals).
To realise this set of business goals, the network requires a governance model ideally realised by policies. Those policies can be low-level configuration parameters, mid-level strategies or high- level intent. If a mix of parameters, strategies and intent is used, it also requires means to translate between them, top-down for realisation and bottom-up for action/benefit as well as validation purposes.
The self-governed network and the governance model must address multiple, diverse constituencies. They are furthermore impacted by non-functional requirements, such as social and regulatory policies. The human intervention should be kept at a minimum focusing on oversight rather than active engagement in control and management activities. Classic self-* capabilities (self-healing, self-configuration, self-maintenance, etc.) are benefits of such as 5G network.
The analysis shows that the RINA theory and architecture not only offers enormous opportunity for flexible differentiated policy management – under multiple policy options – but can also be used to address OAM common problems in a fundamental way that is far less cumbersome than in many of today’s networks.
Section 4.4 of the ARCFIRE deliverable D2.2, “5G OAM: Autonomic Operation & Management and RINA” [page 115], takes this discussion to a deep level by looking at 5GPPP’s view on OAM, which is intrinsically autonomic. From this base D2.2 looks into what autonomic OAM actually means and from there we provide an understanding on how RINA and ARCFIRE relate to autonomic OAM (or provides solution for it). For this reason we discuss one of the autonomic control loops (used for all parts of OAM) defined by the industry and discuss RINA’s relationship to it.
D2.2 explores models, catalogues, and solution sets, and shows a 5G OAM policy definition inspired by RINA with a full explanation of the difference between policy and mechanism with some examples for policies and a final set of policy patterns used in the ARCFIRE experiments.
If you want to see common operations and maintenance problems addressed in a fundamental way, you’ve got to read ARCFIRE D2.2.