Paving the Way Toward Software-Defined Infrastructures

Antonio Manzalini, Chair, IEEE SDN Initiative

SDN and NFV are facets of an overall systemic transformation called “softwarization” which is steering the evolution not only of Networks, but also Service Platforms (e.g., Cloud and Edge Computing architectures) and future terminals, machines and smart things.

Software-Defined Networks (SDN) and Network Functions Virtualization (NFV) are two paradigms that have been known for the past couple of decades, even if with different names (e.g., Active Networking, Programmable Networks, etc): basically they are about the separation of hardware from software and IT virtualization. Importantly, today, these paradigms are garnering a new dramatic attention due to the current techno-economic drivers (e.g., pervasive diffusion of ultra-broadband, increasing performance of chipsets and IT hardware, growing availability of open source software) which are making SDN and NFV exploitable and sustainable in telecommunications.

SDN is fundamentally based on the separation of the software from the hardware. For example, the separation of the control plane of a router from the packet forwarding function in the data plane. In principle, this is applicable to any node of a network such as a switch, a router or even transmission equipment. Another key characteristic of SDN is the possibility of executing software control instructions outside of the traditional specialized network elements, e.g. in standard IT hardware. This means also that a data center filled with general purpose IT servers could execute the control of remote specialized boxes or even network management. NFV regards the virtualization of network functions (e.g., middle-boxes of telecommunications networks) and their dynamic allocation and execution on (almost) general purpose hardware (e.g., x86). It is clear that SDN and NFV are not directly dependent, but they are mutually beneficial: in fact, when they deployed together, they are amplifying the innovation impact on a telecommunications infrastructure.

But we argued that SDN and NFV are more than that: they are two of the facets of an overall systemic transformation called “softwarization” which is steering the evolution not only of Networks, but also Service Platforms (e.g., Cloud and Edge Computing architectures) and future terminals, machines and smart things. Service providers and network operators embracing this vision will see cost reductions, improved operations efficiency and dramatically reduced time to market when providing current and future Information Communications Technology (ICT) services (e.g., immersive communications, artificially intelligent avatars, advanced social networking). But this not enough: Software-Defined Infrastructures (SDIs) should be pervasive and flexible enough to enable radically new service paradigms such as anything-as-a-service: then smart devices, machines, drones and robots will become the terminals of the future, just like “nervous terminations” embedded into the Digital Society. This will mean huge impacts in industrial and agricultural automation, in improving effectiveness in public processes, in saving energy, reducing pollution, etc.

Two main strategies are being adopted for exploiting the “softwarization” in telecommunications. One (called Red Ocean) is based on a relatively slow innovation cycle and it is aiming at a smooth evolution of current legacy infrastructures towards SDN/NFV. This approach runs the risk of being delayed by the standardization of interfaces with legacy systems as well as by the need of updating current Operation and Business Support Systems (OSS and BSS) in order to cope with SDN-NFV advanced features. The other strategy (called Blue Ocean) is based on an open and fast innovation cycle that exploits a parallel SDI in order to create new services markets and ecosystems that go far beyond current competition rules. In this case, rather than getting delayed by trying to integrate new technologies with legacy systems, innovation is disruptively adopted through a “softwarization sandbox;” eventually the transitions from legacy to SDI will be gradual and on a service-by-service basis. In some cases, these two strategies may co-exist, with open source software likely to be the common denominator.

Telecommunications and ICT ecosystems will be profoundly reshaped by this SDI transformation. This recent research report predicts software will disrupt more and more industries. In the meantime the transition to SDI comes with a number of challenges:

• Defining a reference functional model/architecture for SDI;
• Automation of operations processes (as current OSS/BSS cannot easily scale);
• Interoperability between “softwarization sandbox” and legacy infrastructures;
• Best practice, specifications and methods for performance testing and validations;
• Designing security for SDI;
• Defining new biz models and addressing regulations aspects;
• Educating about the change of culture being introduced by “softwarization”;
• Anticipating the needs and creating the conditions for a pre-industrial adoption of “softwarization of telecommunications” through experiments and proof of concept.

Ultimately SDN and NFV will not be just another technological upgrade of telecommunication infrastructures; “softwarization” will be a systemic transformation of several other industries, and much more intertwined with socio-economic sustainability and policy regulations. Research, innovation and exploitation of “softwarization” will require global coherent efforts worldwide which are essential for creating a new socio-economic development for the Digital Society.

Antonio Manzalini received the M. Sc. Degree in Electronic Engineering from the Politecnico of Turin. In 1990 he joined CSELT, which then became Telecom Italia Lab. He started his activities on research and development of technologies and architectures for future optical transport networks: in this context, he chaired ITU-T Questions contributing to the development of several ITU-T Recommendations on transport networks (e.g., SDH, OTN). He has covered several leading roles in European Commission funded projects on future networks. He served as TPC member of several IEEE Conference, and recently he was co-General Chair of EuCNC2014. His results are published in more than 110 papers. He is currently joining the Strategy and Innovation Dept. of Telecom Italia (Future Centre) addressing strategic scenarios and innovation activities mainly related to Software Defined Networks, Network Function Virtualization, Internet of Things and 5G. He is Chair of the IEEE SDN initiative.