In the traditional way of computing availability, the measures mean time between failure and mean time to restore service are employed as parameters. However, for a 3rd Generation Partnership Project (3GPP) fifth-generation (5G) as a service consumer network (that encompasses a cloud-based deployment which is operated by a vendor), any metric that is exposed to a service provider (SP) needs to be refined to indicate the availability of the different services without being improperly influenced by hardware and software failure times. Techniques are presented herein that support the development of an overall network availability metric. Aspects of the presented techniques support the development of an availability measure for each network function (NF) that resides in a network, the determination of all of the valid selection paths in a network (where such a path may be expressed as a graph comprising a set of nodes indicating, end-to-end, the different network nodes that are selected for completing a Packet Data Unit (PDU) session), and the combination of the various (NF-specific) dependent measures to create an overall availability metric for the network. Importantly, the presented mechanism may be extended to private 5G as a service network and Internet of things (IoT) as a service network as well.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.