Future digital networks will be associated with high speed communication, on-demand services and dynamic network orchestration. Current Operation Support Systems (OSS) will struggle in the future digital ecosystem.
The future digital networks are being designed to make digital business a success for the communication service providers and to offer competitive, advanced digital services to their customers. The networks of the future will be predominantly cloud based (virtualized), managing higher volumes of data at higher speeds and ensuring reliable connections between sensored devices. This means that new management techniques have to be incorporated for managing the networks, so that they deliver the expected dynamicity, speed and scale. This will require build-up of the capabilities of the underlying Operation Support System, especially reliability and real-time orchestration.
First of all the digital OSS will need to break the silos of legacy OSS. Far more interworking and interoperation between the OSS layers is needed than ever before. These steps are key to deliver on the promise of ‘Fast, On-demand and Dynamic’, which defines digital networks. Other challenges in stepping up the existing OSS systems include transforming the reactive nature of NOC/SOC to a more proactive one that focuses on early problem resolution. This also means getting rid of well-established but laborious operational processes and inter-departmental hand-offs, which have resulted in slowing down the root cause analyses and corrective actions. To make services agile and their management dynamic, OSS will also need to find more efficient ways of integrating network topology, inventory and assurance processes. And since customers will demand dynamic SLAs, the OSS will have to deal with on-demand capacity configuration and dynamic topology changes too. Analytics to understand the new customers/devices and their behavior will need to keep pace with the service agility of an NFV environment.
On top of this, legacy OSS needs to expand horizontally in order to accommodate data from the new NFV network and, very soon, from the mushrooming IoT network/devices. Most of the IoT devices will serve time-critical applications over IoT networks to serve industries like healthcare, automotive, energy and smart cities. The future digital networks will evolve to carry the rapidly exchanged information, and because of low human interaction, the feedback, reliability and availability ask of the IoT networks will be high. Because of the nature of its applications, IoT will demand high levels of SLAs and policy control. As a result, parts or layers of the network will need to be dedicated to such customers under stringent SLAs in order to guarantee QoS. Typically, the acceptable level of service assurance for the above-mentioned critical IoT applications is 100%.
Many CSPs are also thinking of extending their business to become IoT managed service providers offering high quality, high-reliability networks to deliver IoT Service Assurance, IoT data reliability, IoT service analytics, etc. To make this a reality, the OSS will need to expand significantly to deliver new business rules, service models and mechanisms of device control. To give an idea of the scope of the challenge, OSS systems will be expected to crunch billions of disparate events and data streams pushed through the new network elements. This will force the OSS to build new levels of scalability, real-time capability and reliability.
One would imagine that the current capabilities of Performance Management, Fault Management, Service Quality Management, Customer Experience Management, Inventory Management, Analytics, Automation, etc. would work across physical and virtualized networks. In their current avatars, they would work extremely inefficiently - enough to make the growth of NFV based digital business very slow and painful.
During the disruptive transitions to NFV and IoT, which will last anywhere between 2 to 5 years for most CSPs, network elements will steadily virtualize, capacity of the virtualized network will be constantly stretched, while the capacity of the physical network will shrink. Balancing and monitoring capacity across the physical and virtualized parts of the hybrid network will be critical. The digital OSS system will need to straddle the transitioning networks for a long time, to help plan new investments and optimize capex.
OSS for the future digital network will no longer be simply monitoring, reporting and conducting RCA. It is expected to provide predictive trends of the network, services and customers. For this, OSS will need to employ complex mathematical prediction models to map customer behavior, network investment and revenue potential. It also has to evolve to feed a zero-touch NOC/SOC, which will be fully automated, with closed-loop corrective actions using complex algorithms and machine learning.
In short, CSPs need to evolve not just their networks to the next level, but also introduce intelligent techniques of listening and reacting to the digital network. This can begin by breaking the traditional OSS silos and integrating their functionalities in innovative, clever ways.