End-to-end orchestration gathers momentum as 5G scales up

The orchestration of telco operations has become a pressing issue as 5G starts to scale and new service models come into play, as TM Forum's principal analyst, Dean Ramsay, explores in this extract from TM Forum's recent report Key Themes in 2022.

The concept of orchestration of telco operations came to prominence around the time that network functions virtualization (NFV) and software-defined networking (SDN) started to gain traction as the new way to build networks.

One of the original ideas around management and orchestration was that orchestrators would provide a master control which connected service-layer concerns with network-layer capabilities. In doing this, it was hoped that communications service providers (CSPs) could use technology to force change in the industry and make operations truly service-centric and customer-focused.

This concept became one of the central tenets of 5G operations and the TM Forum Open Digital Architecture (ODA). The problem has been that in many deployments CSPs have been forced to introduce a proliferation of orchestrators to mesh the multivendor operational support system (OSS) environment, stacking the control of service orchestration in a complex web. The complexity of this situation is untenable going forward, and there is an increasing amount of dialogue within TM Forum member collaboration groups about how best to address this situation, and indeed how to address it by the end of this year.

5G scales up

In 2022 the 5G story is once again gathering pace, as we move to 5G core networks going live and the beginning of standalone (SA) 5G. With this progress, 5G network slicing is starting to be a commercial reality, but the key component for CSPs to provide the automation needed in both service-layer and network-layer operations is orchestration.

Scaling non-standalone 5G operations – and monetizing the service models based on that network and IT framework – will rely entirely on a CSP’s ability to orchestrate processes from the customer order touchpoint through fulfillment to network activation and on to post-activation activities, like assurance and revenue management. This is the challenge that CSPs are now acutely aware of, and they are working with their technology partners to make 2022 the year that end-to-end orchestration is streamlined and optimized.

One key part of this effort is in centralizing the role of the service catalog in the OSS. Swedish network operator Telia is using the TM Forum Open Digital Architecture to automate its fulfillment process for products that rely on physical and virtual networks. This has been challenging because the company operates in six countries where each operating company had its own catalog deployment with no common platform, making it impossible to leverage scale, commonality or standards. Read about how Telia is approaching these challenges.

New service models

The majority of new service models and business opportunities that CSPs have identified for revenue growth in the next decade have a new technological element, which was previously missing or immature in the LTE era. For example, 5G cloud gaming partnerships would see CSPs in a value chain with games developers, hyperscale cloud providers, device manufacturers, and an ecosystem of cloud-enabled specialist IT companies. As the compute and hosting for the game would exist at the network edge, a multiaccess edge computing (MEC) function in the 5G network is essential to provide a near-zero latency connection to the end user.

The platform-based cloud ecosystem, which links the user’s instance of the game to other geographically diverse users, is also a relatively new concept for CSPs to deal with.

Beyond these new technologies there are more familiar activities the ecosystem of partners must carry out, such as order management, service delivery and billing, but in a fully automated way through their shared platform model. The role for service orchestration here is made more complex by this requirement to seamlessly organize workflows and automated processes beyond the usual limits of the CSP’s own IT end-to-end orchestration gathers momentum as 5G scales up and out into a multi-cloud world.

The end-to-end view has extended out past the operator’s usual demarcation points, but CSPs must address this challenge if they are to prosper in the value chains of new digital service business models like the gaming example given here.

Showing intent

Intent-based networking is a conceptual trend which really gathered pace at the end of last year. It focuses on a CSP’s ability to translate its business goals or customer demands into policy sets which can be actioned by dynamic network and service functions in a highly automated way. This is the reverse concept of traditional telco operations where we look at what the network is capable of and then design service models around that functionality. Intent puts the customer and the CSP’s business goals at the center of service model design.

This idea was something that became popular a decade ago when CSPs started virtualizing their networks, introducing NFV and SDN. The service orchestration piece was seen as a way of providing master management and control of the automated workflows down through the IT stacks into network orchestration and the groups of controllers, with element and network management systems (EMS and NMS) managing the network below. During this idealistic phase of conception, the end-to-end orchestrated view looked like the answer to classic telco problems, but it proved to be very difficult to implement in complex and messy multivendor, multi-domain operations.

In 2022 we now see this idea coming around again under the momentum of intent, but with the flexibility and power of the telco cloud and mass Open API adoption to support it.

AI and data management

Another key indicator that this current technology cycle may have all the right elements to make end-to-end management and orchestration work, is the growing maturity of artificial intelligence (AI) and the use of complex data in telecoms operations.

Closed loop operations in the OSS is now becoming a reality for some CSPs, and the next phase is introducing some human-like intelligent decision-making into the processes, at scale, by using AI.

The importance of this step in the ongoing development of orchestration cannot be overstated, as it will remove a great deal of the need for manual remedial work from the process flows which fulfill new services or make dynamic changes to existing services.

Read our recent report Autonomous networks: exploring the evolution from level 0 to level 5 for a detailed explanation of AI’s impact on service operations.