The white paper captures 10 years’ experience of commercialising SON (self-organising networks) in over 18 million small cell deployments.
The ‘SCF233 Small Cell SON and Orchestration from 4G to 5G’ paper details implementation recommendations for key 5G technology trends including disaggregation, virtualisation, private networks and analytics with artificial intelligence and machine learning (AI/ML).
The paper builds on the small cell industry’s experience in establishing interoperable plug-and-play small cells for 3G and 4G. The collaborative paper was produced by members AirHop, CommScope, Node-H, Nokia, Parallel Wireless, Reliance Jio and Samsung Electronics.
Explaining the reasons behind producing the paper, SCF said that ambitions for 5G to address future connectivity needs will see small cells deployed at scale in enterprise and industry verticals. This points to the need for pervasive end-to-end automation to ensure network operation remains feasible and affordable. Furthermore, advances in artificial intelligence (AI) and machine learning (ML) have been shown to optimise dense het-nets in ways that are beyond human capability.
Dr. Prabhakar Chitrapu, Chair of Small Cell Forum, said: “With over a decade’s experience of commercialising large-scale plug-and-play small cell deployments, this paper presents SCF’s recommendations of how SON and automation need to be implemented for the broader end-to-end and AI-enabled automated vision for 5G.
“Open and interoperable management and orchestration are a pre-requisite to realize these ambitions. Standards are an important first step, but further industry effort is now needed to prioritise scenarios, architectures and features for use in multi-vendor interoperability testing,” said Chitrapu.
“It is evident that with increasing complexity and dimensionality of networks as we move towards the 5G-Era, self-organisation and automation are key for network operators,” said CommScope’s Balaji Raghotaman, who led the project.
“As the industry moves towards a virtualised environment with off-the-shelf hardware, small cells are also moving in that direction. This brings opportunities and challenges, including the need for SCF to develop lean and scalable management models for small cell scenarios, which can complement other industry efforts defining feature-rich but information-heavy macro cell counterparts.”
Following these recommendations, SCF members have initiated a number of work items which look both at the underlying management platforms, as well as future AI/ML applications that open and interoperable management and orchestration will enable.
