5G networks will change the way we live and experience the world, with nearly instantaneous speeds, new services such as Mixed Reality (MR), Augmented Reality (AR) and Virtual Reality (VR), and most importantly an expanding ecosystem of enterprise solutions, involving network services that can be scaled and right-sized for wholesale business-to-business-to-customer or enterprise (B2B2x) solutions.
Despite the hype, however, the first 5G services won’t arrive with much fanfare: changes will be rolled out slowly and some customers might not even notice the initial differences.
It might be tempting to think that 5G changes are incremental, and while major network investments are required, business systems won’t require big changes. Nothing could be further from the truth. In addition to changing our lives and habits, 5G will revolutionize business models for consumer devices, smart cities, factories, agriculture, and venues. Not only will 5G deliver a massive increase in the volume of transactions, it will involve a massive variety of transactions — which raises a fundamental question: how should customers be charged for these services?
Traditional telco billing/charging systems were built for a simpler world, where services are often measured in minutes and megabytes. What happens when new pricing schemes are needed for millions of users making small transactions with thousands of different services? How can service providers spur interactions and automate new functions harnessing rich analytics across a breadth of new network functions? If a charging system isn’t built for a 5G environment, it will fail in all these aspects, which means another year of flat revenue for communications service providers (CSPs) and a few more steps along the march toward being a dumb pipe.
5G isn’t just another “bigger and faster” network. Yes, 5G networks will offer exceptional speed and scalability, but it’s not just a bigger bit pipe. The next generation network is positioned to empower CSPs as digital service providers, harnessing a flexible, automated, cloud-based architecture to launch new services quickly and employ the same web-scale efficiencies as digital leaders like Amazon, Facebook and Google. This requires a cloud native approach.
Cloud native is a specific set of architectural requirements defined by the Cloud Native Computing Foundation (CNCF) that includes the use of containers, microservices, dynamic orchestration, and the separation of stateful and stateless services, among other things. Updating an existing legacy solution to run in a virtualized environment is not cloud native. It would be more appropriate to call this a cloud tourist, as it can exist in a cloud environment, but isn’t adapted to it. Cloud tourists will still face many of the same performance and scalability limitations in a cloud environment as their non-virtualized counterparts.
Cloud native solutions are optimized for distributed environments, regardless of the platform infrastructure. Well-designed applications running in distributed systems are central to the promise of 5G and will enable an evolving ecosystem of massive new IoT devices, slices for enterprise and new business models.
Software solutions like these run natively in containers (e.g., Docker) with container orchestration (e.g., Kubernetes), are designed as loosely coupled microservices that use shared resources, use lightweight APIs to collaborate with other solutions, and handle elastic and resilient deployments as only a truly distributed and dependable architecture can.
5G is revolutionizing the kinds of digital transactions that will happen over many new emerging devices. These could be billions of tiny transactions from smart meters installed in our homes or in cities. Or they may be dozens of transactions processed through our smartphones as instant transactions or digital shopping carts. 5G will not only bring an increase in the volume of transactions; it will involve a massive variety of transactions.
Some 5G transactions will be stateless, such as “one and done” transactions like buying a movie. Other transactions will be stateful — for example, paying a per-minute charge for a live video chat with a healthcare professional. Some may require extremely low latency (e.g., virtual reality experiences), while others may need to be scaled up and torn down quickly (e.g., augmented reality services at a live event).
Because of the demands of differentiated services, 5G networks are fundamentally different by design. With 4G, monetizing the network means charging more for data and video services. In 5G, however, CSPs will need to be creative. They’ll need to slice their network to provide very different experiences for their customers: one slice for low-cost, high-latency IoT traffic, another slice for high-cost, low-latency emergency response services. In addition, CSPs will need to be able to create, deploy and tear down services quickly if they expect to create value in an ecosystem involving their agile, over-the-top (OTT), digital-native competitors.
Charging and billing systems must adapt to this new world. As new network slices are created, new billing and charging services will need to be created to support those slices. These charging services should be designed to scale up and down seamlessly, move out of the core network and into the network’s edge for lower latency, apply different policies for security and quality, and a myriad of other considerations. In short, charging and billing needs to behave like a cloud native so it can be easily integrated into and meet the performance requirements of 5G services.
The business models for 5G will be radically different than those for 4G services. New services will be more immediate and transparent, providing incremental value via “networks on demand” and through tailored services, with nuance and scale not available in legacy networks. These distinctions will impact both enterprise and consumer pricing as 5G matures. In turn, solutions are needed that are ideally suited to support billions of stateful transactions at the scale and speed that 5G services require.
Design principles for 5G charging systems must evolve likewise, both with respect to application and database design. In order to understand why a new database may be needed for 5G scenarios, one must consider where other databases fail. To begin with, telecommunications networks are very different than most computer networks. For example, most networks can sustain some level of packet loss, re-sending the lost packets at a later point without losing the integrity of the session. Any kind of loss is intolerable in real-time revenue bearing networks, however.
Similarly, most computer networks consist of stateless transactions, in which the state of the session can be lost without impacting the transaction itself. A stateful transaction is a different story. It requires that the state of the session be maintained with resilience, so data is not lost when failures inevitably occur.
To ensure stateful transactions in a 5G world, one needs a database with exceptional elasticity and resilience. The majority of industry databases use eventual consistency or asynchronous replication, neither of which provides the accuracy or performance required. An option involves an all-primary data store architecture allowing each transaction to be simultaneously committed to multiple primaries. If a transaction does fail, it will still be maintained by other primaries and may be accessed seamlessly to provide service continuity. 100% accuracy and near instant performance are both required when charging and billing 5G services in a modern environment.
5G charging systems should also harness the full power of open source tools from the Cloud Native Computing Foundation to achieve much greater scale in distributed systems. However, careful design choices must ensure both the performance and consistency of telco transactions in such an environment. Without this level of performance, the charging system cannot act as a reliable management service in 5G networks, coordinating millions of valuable transactions every minute, delivered instantaneously and seamlessly to customers, and enabling new streams of revenue for the operator.
5G offers a world of opportunity for CSPs, particularly in the way that network services will be created, deployed and most importantly, monetized. Where the revenue focus in the past has been on direct-to-consumer services and, to some extent, direct-to-enterprise services, a new revenue focus for 5G will ultimately be on B2B2x (consumer or enterprise) services. In this new landscape, CSPs will no longer simply be technology providers, but business partners that help other businesses innovate, create and deploy exciting new services.
A legacy billing and charging system has no place in a 5G network. It can slow down revenue streams, create frustrating customer experiences and ultimately compromise CSPs’ ability to compete with native digital service providers. To win in a 5G world, a cloud native charging system delivering scalability, performance and reliability is what CSPs need to successfully stand up and monetize new 5G services. It is a key investment that will pay off for years to come.
NOTE: This article first appeared in Telecoms.com, March 2, 2020.