Making the move to ‘zero-touch’ service and network operations
End-to-end network and service orchestration and automation are no longer just ‘nice-to-haves’ – they are essential if telcos are to realize the full potential of their networks and thrive in an increasingly competitive marketplace.
Digitization and automation have been the core focus for communication service providers around the globe for the better part of the past decade. The move towards the automation of next-generation networks has received further impetus by two more recent developments.
COVID-19 has been an unexpected accelerant, driving the need for networks to operate with greater autonomy to avoid face-to-face contact, for example by having to send engineers out for repairs.
At the same time, 5G deployments have started in earnest, and a new generation of services is expected to emerge, creating an intricate web of moving parts. 5G standalone networks will be largely based on cloud-native and SDN (software-defined networks) concepts, and in setting them up CSPs have an opportunity to bring in automation natively – without the complications of having to bolt advanced automation onto existing networks.
That said, 5G networks will only add to the multiheaded hydra of physical and virtual networks, multi-vendor environments, different cloud flavors, and a vast number of domains that are all currently managed separately.
How can CSPs ‘stitch’ all these together and create an environment that’s not only efficient and cost-effective to run, but also lends itself to fast-paced service innovation, which will be vital in the battle against cloud and over-the-top providers? This is where end-to-end network and service orchestration comes into its own – and is moving us closer to autonomous self-healing operations.
A single view of the network
Currently, each network domain is managed separately with its own OSS and orchestration tool-sets – in many cases with some partial automation, but little else. The siloed, stove-piped approach taken to date – all for good reason in the days of physical networks – means there is little to no overarching visibility of the network and its support systems. This makes it hard to optimize how they work with each other.
Stovepipes also hamper network openness and with that, the advantages of harnessing the collective strengths of multi-vendor, multi-domain environments and multiple network functions to deliver innovative services.
Without end-to-end network and service orchestration, it is impossible to obtain a single view of the network and manage it centrally. And without central management, providing services that combine capabilities across multiple domains and environments becomes virtually unfeasible. What is more, end-to-end orchestration gives CSPs the adaptability and responsiveness needed to react more proactively to customer demands and market opportunities.
You could say that orchestration plays a critical role like a PC operating system, linking all parts of the network and its support systems together.
Toward closed-loop operations
Longer-term, the aim will be to move from centralized management to closed-loop operations. Here, automation will be so highly developed that it monitors networks and services, and acts on any faults, congestion or capacity needs with minimal (if any) human intervention (zero touch).
Ultimately, closed loops will evolve to autonomous networks that can heal and adjust themselves based on the cues they collect and analyze from across all network activity.
In 5G environments, for example, this might include the ability to monitor many network slices, tending to faults, or adjusting capacity needs according to network activity and the services levels that need to be met. Actions will be taken independently based on pre-defined thresholds and rules on the one hand, and machine learning and AI on the other.
Another example is executing on enterprise customer bandwidth requests automatically, perhaps to accommodate an increase in home working, such as many companies have experienced during the pandemic.
Seizing the opportunity
While there is an investment to be made into end-to-end automation and orchestration − from the customer to the network and back − the advantages outweigh the capital requirements by far.
The customer experience benefits from end-to-end automated orchestration – services can be provisioned more quickly, and faults detected and remedied remotely, without an extended wait.
With network operations one of the costliest and most labor-intensive domains in OPEX terms, closed-loop operations can bring about dramatic efficiencies for service providers.
End-to-end network and service orchestration will also provide a path for CSPs to overcome legacy monetization strategies based on volume, time, or events. Instead, they can, for example, offer enterprise customers dedicated network slices for different services, each with their required latency level as well as differentiated QoS and security. And they can charge accordingly, ensuring a good return on their investments.
Finally, as business and customer services will increasingly be delivered by an ecosystem of interconnected partners, the ability to open up networks to facilitate partner access will also pay off. This will help service customer needs in a more differentiated manner, thereby tapping into new revenue streams.
A healthy position for growth
With the arrival of 5G and the need to focus on innovation rather than the day-to-day running of the network, introducing end-to-end orchestration and automation is no longer just a ‘nice-to-have’ – it’s essential.
CSPs must take this step to survive and thrive in the intensely competitive, increasingly fragmented marketplace for communications services.
And 5G presents a perfect starting point from which they can expand their end-to-end automation across the entire operation over time. Beyond the obvious business and customer benefits, the work will place CSPs in a healthy position for the future.
About the author: Eyal Shaked is General Manager, Amdocs Open Network Product Business Unit
