By: Cyril Doussau

Network functions virtualization (NFV) and software-defined networking (SDN) characterize the telecommunications industry’s path to the future, enabling CSPs to offer a bevy of new value-added services at a time when prices for standard connectivity are beginning to bottom out. Among these offerings are commercial-grade network-as-a-service (NaaS) products that can deliver fast, on-demand service activation, all in the pursuit of delivering the best possible customer experience.

But this switch to a fully virtualized, software-defined network, enabling the digital telecommunication industry of the future, isn’t going to happen overnight, as full adoption is still projected to be four or five years out.

As CSPs launch “as-a-service” network products, they should not lose sight of customers’ QoS

As we know, service-level agreements (SLAs) bind CSPs to delivering a consistent standard of service over a given term. A CSP cannot upend an existing network architecture – even if the end goal is an improved suite of services – without taking into consideration how this evolution effects existing SLAs and quality of service. 

Performance visibility and SLA monitoring is a prerequisite to any network transformation, especially when CSPs will have to offer new innovative services in tandem with the ones they are already delivering to customers. As next generation services will be instantiated dynamically, OSS solutions will have to align in real-time with changes happening within the service delivery chain. This means that, to realize the benefits of NFV and SDN, a fully orchestrated service assurance platform will be required. 

Network transformations require service performance orchestration

Orchestrated service assurance is an especially important solution when it comes to offering the value-added services that CSPs need to retain and grow their customer base. For instance, without having complete visibility and understanding of current end-to-end network performance, how can a CSP offer network connectivity with specific SLAs through a self-service portal? This lack of insight is equivalent to a box office selling tickets to a show without having any idea whether the venue has met capacity or not, inevitably resulting in a slew of paying customers not enjoying the experience they paid for. 

Similar visibility is also crucial for those leveraging network controllers to automatically configure virtual and physical functions. How, for example, can an SDN controller reconfigure a network function without an understanding of how these changes will impact a customer’s QoS? While automatic configurations may help cut operational expenditures, controllers could inadvertently ignore how they affect customer service. This would be like designing a building in a way that cuts costs and completes the project as quickly as possible, but neglecting to consider critical safety codes in the process.