Digital transformation offers new opportunities for communications service providers (CSPs) to flexibly deliver new network services over software-defined, cloud networks, such as 5G. However, this new service arena is complex to manage and involves orchestration and management of services and networks across multiple domains. The scale and scope of the new service environment is way too complex to be handled by anything other than automated systems. An OSS evolution including end-to-end capabilities is needed if CSPs are to make their automated dreams a reality and monetise their new networks.
Eyal Shaked is general manager of Intelligent OSS at Amdocs, a provider of software and services to the communications and media industry. He is responsible for leading the business from strategy, product roadmap and R&D to deployment and professional services.
A 30-year industry veteran with a strong track record leading teams to deploy complex systems, Shaked was previously managing director of Coral Group, a venture capital firm. Prior to that, he held various roles that included chief operating officer at Playtika, as well as various executive positions at ECI Telecom and ultimately, general manager of their Network Solutions division. Shaked worked on high-scale projects for major service providers including Deutsche Telekom, BT, France Telecom and Vodafone.
Shaked holds an M.Sc. in electrical engineering and computer science from Tel Aviv University and a B.Sc. in electrical engineering from the Technion – Israel Institute of Technology.
Now, he tells George Malim, the managing editor of VanillaPlus, it’s time for CSPs to bring together their existing systems with next generation intelligent OSS tools to enable automation and the new service arena that will transform CSP revenue generation.
George Malim: Automation has always been essential for enabling the efficient CSP. What are some of the new drivers and requirements for network and service automation?
Eyal Shaked: The major change in the network is the move towards enhanced automation, while evolving from siloed to cross-domain systems – with physical and virtual infrastructure blended together. At the same time, cloud and the network are being mixed and we are expecting implementations of distributed, micro-pool edge clouds to support Open RAN (ORAN) with virtual network functions and cloud-native network functions. We are already seeing adoption of both on-premise and public cloud driving complexity and all of these steps are being taken in order to provide better services to the customer. However, the vision can’t be realised if it is not holistic. Automation needs to be end-to-end, it can’t be done in siloes if the new, automated era is to become a reality.
This new era demands an OSS evolution, a next generation intelligent OSS. The network domain is more programmable, more open and more software-driven but this doesn’t mean that other technologies are going away – new capabilities and technologies are being added on top. Containerisation, for example, is being added in parallel to use of virtual and traditional, physical network functions and this all needs to be automated to deliver efficiency. For the industry this is a big change in pace because upgrades used to be an incremental change but now involve a dramatic evolution across multiple domains in parallel. We’ve focused our next generation OSS strategy to be ready for the future and to be able to connect it all together and enable hyper automation. It’s the only way CSPs can move to the new era.
GM: How will CSPs reduce their opex even as complexity increases?
ES: The need to operate efficiently drives automation and concepts such as the dark network operations centre (NOC) meet the industry’s needs to reduce opex. The complexities I have just described exist and you either manage them by increasing the number of people that manually operate the network or address it with them with a higher degree of automation. The arrival of artificial intelligence (AI) and machine learning (ML) delivers greater capabilities to anticipate problems before they affect the service and these are therefore substantial enablers of automation.
The combination of better assurance of the network to understand the fault and the performance together with orchestration to identify and secure and operate the network and heal it before it fails is a compelling improvement. Achieving this starts with the data that provides the fuel for AI. This end-to-end network information from multiple sources needs to be understood in order to perform root cause analysis. This is not just about the network, it’s also about the business and the data needs to reveal the business risk, not just the network performance impact. For example, a service provider would like to know that a group of customers, VIPs for example, are experiencing service degradation. Once they know this, they can mitigate customer complaints in advance, while restoring the service to its committed service level agreement (SLA) in parallel.
To achieve that closed loop which powers the dark NOC, you need a comprehensive management of the service and network lifecycle and this stems from an accurate catalogue. If you don’t have that – you can’t deliver intent-based orchestration and apply it to particular use cases. If you connect the catalogue and the operating model to cross-domain inventory, that gives you the most accurate view, enabling you to design, deliver and constantly maintain the customer’s intent.
This approach to operations is central to the Amdocs Intelligent Networking Suite, enabling intent-driven, closed-loop operations with a cloud-native suite of pre-integrated products.
GM: We’ve talked about operational efficiency gains but what impact does automation have on revenue generation?
ES: The impact is major. We work closely with the market leading consulting firms that seek our advice. We see a substantial transformation from legacy OSS to next generation OSS and this in itself opens up new revenue opportunities, enabling CSPs to deliver and maintain new network currencies so they can be monetised. If a CSP wants to sell a guaranteed performance level for a particular service, it must have the capability to control and manage the network to deliver the service intent. Moreover, CSPs can expose their networks to external parties, allowing them to change the SLA, quality of experience and other aspects in near real-time and according to their contextual need. Enabling these innovative capabilities will generate new revenues for CSPs.
Network management systems today are siloed but the situation needs to change. We are leading the change helping CSPs to evolve towards a single umbrella system to manage, optimise and heal the network in real-time. This is where the market is going and it is being driven by the service the network needs to support. If a CSP wants to provide different quality for different services, it has to be able to allocate dedicated network resources to run the differentiated services, for instance creating different slices in a 5G network.
In terms of revenues, we’re seeing a lot of investment on new 5G networks which is focused on services like IoT, gaming, augmented and virtual reality, or drones. This investment needs to be followed with OSS modernisation to avoid a chicken and egg situation where, if you don’t have automation, the specific services or network capabilities to support them can’t be delivered. Automation is a foundational step; it’s not just about deploying new networks.
Automation of OSS must be business driven and focused on service intent. The system needs to be smart enough to figure out the best way to give the service what it needs from the network and maintain it that way throughout the service lifecycle.
GM: How is Amdocs helping CSPs increase automation and prepare their OSS for the new era?
ES: We are involved in multiple projects and also running a wide range of innovative proof of concept projects.
We have a proven network-as-a-service (NaaS) offering, actually in production at Vodafone Ziggo in the Netherlands in which virtual network functions need to be orchestrated across multiple domains and the network needs to be automated and exposed as-a-service to a business domain.
Our 5G Slice Manager offering is also well received across geographies. In the US, we have launched with a tier-1 operator one of the largest enhanced mobile broadband (eMBB) 5G slice-based services on a 5G standalone network. This is where the focus shifts fundamentally as lots of network slices are implemented across the RAN and orchestrated across dozens of systems. We are helping our customers find the optimal path to progress on, making use of their existing systems but changing what is needed on a case-by-case basis. We draw upon our industry expertise and relationships to go with our clients through this process. The variety of projects we are involved in is wide.
We are also involved with another tier-1 operator in the US in a proof of concept to orchestrate 5G slices across both Ericsson and Nokia network equipment – proving the importance of being truly vendor agnostic.
In Europe, we have successfully run a proof of concept with A1 Telekom Austria including a very advanced set of business cases that involve dynamic network slicing with closed loop operations, and orchestrating both the cloud and edge – while powering new monetisation triggers from the network.
We are at a very interesting step in the journey where CSPs are learning how to design, create, activate and maintain network slices. Everybody talks about it but getting real and having a system to automatically build network slices is a big challenge. We are doing it already, developing the technology that allows the CSP to efficiently manage network slices, working with orchestration and designers to get the right steps implemented.
GM: There seems to be a fusion of old meets new in OSS in order to enable the automated era. How has Amdocs enabled traditional OSS to interoperate with new orchestration and other systems?
ES: Very few CSPs are starting from zero with a greenfield approach so it’s essential that existing systems are incorporated. In addition, many of the functions of traditional OSS are still needed in the automated era. We’ve assembled a wide OSS portfolio to support the journey to automation and that includes our expertise across traditional OSS.
A great example is inventory which traditionally has been a static and siloed environment that is now evolving to be a vital enabler of automation. As we have discussed, the network is evolving to become software-defined with functions in the cloud operating across multiple dimensions and, in order to deal with that, we have to bring existing, legacy technology together with new approaches. That means bringing network equipment such as IP and WDM hardware together with cloud and virtual network functions into one multi-domain, multi-layer inventory. That unified inventory is essential to understand what happens in the network.
This isn’t just about enabling visibility of what a CSP has got, inventory is now a critical data source across multiple domains to enable automated fulfilment and closed loop operations. It provides the data to plan the network, to heal the network and to operate it in real-time to deliver and maintain the service required.
To help achieve this, we’ve built an award-winning orchestration and inventory on the same platform so CSPs can have just one platform that can take different capabilities and integrate them in real-time. That encompasses both the data on what needs to be done to operate the network and what is the current state of the network. We are exploring use cases and how they are best implemented and the industry realises that this is the way to make the automation vision a reality.
GM: How essential to the future of telecoms is automation of OSS?
ES: The bottom line is that we, and our customers, see automation as key to the industry remaining competitive and to supporting 5G-era networks plus emerging new services. We have taken our experience and are using it to help our customers to find optimal ways to make use of their existing services while adding our new offerings to drive their automation to the next level. This is a complex journey that we are going on together and we keep assessing, optimising and further automating to keep projects on track.
The end goal is to get everything working together and making it work at scale in real-time. The ability to integrate across domains, adhere to standards and streamline the complexity is what enables not only greater automation of OSS but ultimately the automation of the network to deliver new services on-demand, and monetise them in completely new ways.
Comment on this article below or via X: @VanillaPlus OR @jcvplus
