Wireless networking, around the world, is revving towards a total metamorphosis. Newer, scalable, intelligent, cost effective and Open technologies are spreading their wings to alter the existing wireless communications networking landscape, like never before. This phenomenon has become more palpable in the past few months of 2020, when large tracts of our lives and modes of communication have become almost entirely digital.
The virtual takeover
Our workplaces are increasingly virtual, our shopping is virtual, our education is going virtual, our entertainment is predominantly virtual and the delivery of healthcare is going the virtual way. Most importantly, many of our closest connections are sustained through virtual platforms.
This recent shift makes it imperative for builders of these technologies and platforms to make their products and services more intelligent and compelling (for users), while also being more open and feasible, for themselves.
Software Defined Networking (SDN) Technology
In the last decade, SDN Technology that emerged as a technology of choice among many enterprise networks was aimed at making networks more flexible, cost-effective and capable of integration. It helped improve network performance by decoupling the network control and forwarding functions, thereby enabling network controls to be programmable. SDN communicates with applications via an Application Programming Interface (API) and allows users to control virtual resources through control planes, instead of physical infrastructures such as routers and switches. It can virtualise a whole network, as compared to conventional networks.
Open Radio Access Network
The next frontier in making an Open, Virtualised, Software Defined Networking infrastructure is the Radio Access Network (RAN) of mobile networks. With many initiatives such as OpenRAN by Telecom Infra Project (TIP) and ORAN alliance, the movement towards an Open ecosystem is gathering momentum.
Rakuten, a Japanese ecommerce platform company turned mobile network operator led the world by deploying the world’s first completely cloud-based Greenfield mobile network, based on a virtual RAN concept. OpenRAN is being used to Virtualise Radio Access Networks to meet the growing demands that require scalability, increased management efficiencies and profitable operations. The virtualized architecture deconstructs parts of the RAN and allows service providers to gain economies of scale, through shared software-defined functions, while also improving processing efficiency required by latency-sensitive applications.
Machine Learning (ML) & Artificial Intelligence (AI)
In addition to bringing a new dawn to the ecosystem of telecom equipment providers, Virtualized OpenRAN has thrown open a new, more intelligent network infrastructure driven by AI/ML. This was not possible in the earlier aggregated network. With more computing power available in virtualizsd, cloud-based platforms, the scope for adding more intelligence and automation to the network has a huge potential. Now, this is where newer start-ups are bringing in innovation and newer ways to manage and operate – networks more efficiently.
Even before the COVID 19 enforced WFH environment, data traffic was continuously straining the capacity of communication networks. With the current COVID-related increase in online activities, plus the influx of new applications of communications such as the Internet of Things (IoT), wearable devices, autonomous systems and drones with vastly different performance requirements, even more data traffic is being generated. This growth makes it necessary for more intelligent processing, operation and optimisation of communication networks.
ML, the vessel that carries AI, is being incorporated into design, planning and optimisation of future communication networks. These new applications enable intelligent communication designs and also address challenges such as signal detection, sparse signal recovery, channel modelling, network optimisation, resource management, routing, transport protocol design and application/user behaviour analysis. In addition to smart network management, ML will allow future communication networks and their applications to exploit Big Data Analytics and thereby enhance overall network operation.