Traditionally, network convergence refers to the provision of telephony, video, and data services within a single network. Today, this convergence is a given. The network convergence we are looking at today, with the horizon of 2020 and the maturing of 5G, is the provision of a very efficient network (better performance for lower cost), a very fast network (very high data rates, spectrum efficiency, coverage, and low latency), and a converged fibre-wireless network (heterogeneous physical media) all with the same physical and virtual infrastructure. On top of that, the network should support the so called verticals, such as automotive, transportation, healthcare, energy, and manufacturing; each with very different requirements and constraints. Consequently, ARCFIRE has focused on RINA for a 5G converged service provider network, its design and realisation.
5G has matured in a number of areas since the start of the ARCFIRE project. However, the complexity to define what a 5G network is, means that new concepts, architectures, and learnings do appear frequently. ARCFIRE is part of this dynamic environment, which creates a challenge in providing a future-proof and consistent report with added value that we can contribute back to the 5G R&D efforts.
ARCFIRE addresses this challenge by looking at new forces for operators such as disaggregating the radio access network, redefining the core network, and evolving the transport network. ARCFIRE takes two representative studies as basis to identify the facilitators of 5G, the goals set out be various stakeholders, the recommendations relevant to ARCFIRE, and the final challenge in that RINA and ARCFIRE are keen to address: automation.
ARCFIRE has brought forward and discusses an ideal RINA network in 5G. This discussion starts with general design considerations followed by detailed views for the access network, metro aggregation, and the core network. ARCFIRE looks at concrete 5G scenarios (radio access, cloud), autonomic control, and policy-based management (called strategies in RINA). It also provides a first look at advanced ARCFIRE work on RINA-inspired modelling for 5G operation and maintenance, such as a general resource abstraction model and adaptive policies (for higher level management strategies). Each part of this introduces a 5G concept (or scenario or challenge) and then details how RINA in general and ARCFIRE in particular can address it.
ARCFIRE then backs up these discussions with a view on ARCFIRE experiments and how they relate to the converged service provider network design discussed in D22. Such experiments will have immense impact on the networking community and 5G R&D.
