(Image Credit: iStockPhoto/Todor Tsvetkov)
A consortium of vendors, operators, and research institutions has launched the Superfluidity project to design a cloud-native 5G network platform. The initiative aims to deliver a converged virtual network and service infrastructure that spans from the mobile edge to the 5G core and into data centers, enabling operators to deploy and manage services more efficiently and responsively.
Superfluidity targets persistent limitations in current networks, such as long provisioning cycles and the common practice of over-provisioning to handle variable demand. The project also addresses the dependency on rigid, costly hardware and the complexity that arises from three distinct types of heterogeneity: diverse traffic patterns and sources; a wide range of services with different requirements; and multiple access technologies combined with multi-vendor components.
To overcome these challenges, Superfluidity follows a multi-faceted strategy organized around five core principles:
- Flexibility — Decompose network functions and services into small, reusable building blocks. This modular approach lets operators combine primitives to create tailored network services without redesigning the entire system.
- Simplicity — Adopt a cloud-based architecture that eliminates access-specific gateways and integrates varied access technologies into a unified cloud-network fabric. This convergence reduces operational complexity and streamlines management from the core to the edge.
- Agility — Virtualize radio and network processing tasks so that resources can be scaled and orchestrated dynamically to meet changing demand and to support rapid service deployment.
- Portability and viability — Define platform-independent abstractions that enable network functions to run across heterogeneous hardware. At the same time, the architecture permits vendors to use closed platforms or proprietary implementations where needed.
- High performance — Achieve performance beyond current software-only approaches through software acceleration and hardware specialization. The project seeks to make such optimizations transparent to service developers so they can focus on creating innovative services instead of tuning performance.
The consortium brings together major industry players and academic partners. Commercial participants include NEC, Intel, Citrix, Red Hat, and OnApp, while operators involved include BT (British Telecom), PT Inovação e Sistemas, and Telefónica. Research and academic contributors include CNIT, the University of Liège, the Dresden University of Technology, Ben-Gurion University, and the Polytechnic University of Bucharest.
Alcatel-Lucent serves as the technical coordinator for the Superfluidity project, with CNIT acting as the project coordinator. The initiative began on 1 July 2015 with a planned duration of 30 months and was scheduled to complete before 1 February 2018.
What do you think about the Superfluidity project? Let us know in the comments.