5G networks will undeniably enable many extraordinary new services and revenue possibilities for communication service providers (CSPs). To leverage these opportunities to the highest potential, CSPs require a flexible, high-performing packet core that will deliver the low latency and high-speed connections these services demand. That said, there are several areas that CSPs should consider when investing in a packet core.
High performance, flexibility and scalability are fundamental
One of the areas that are the most important for CSPs is high throughput and low latency. A virtual forwarding path (vFP) architecture is critical here, enabling throughput per user equipment in the order of 10Gbps or higher and cutting latency values to as low as 0.3ms.
Flexibility and scalability are other areas that are important to CSPs. To ensure a flexible and scalable network, CSPs will require a packet core solution that provides multi-terabit packet core user planes based on physical network functions and pre-integrated packet core appliances. These will act as a control plane or user plane, with features that support a diverse range of business needs.
Minimizing implementation risks
One of the most common challenges faced by CSPs is how to minimize the risks inherent in implementing a 5G packet core. The network must be deployed regionally for coverage, ensuring low latency, and distributed close to the edge. All of this is to serve a demand that is very often unpredictable.
To minimize this risk, CSPs can stay nimble by deploying the core network on a public cloud, taking advantage of hyperscalers’ infrastructure to rapidly serve a demand. This can be a temporary implementation, i.e., in case the demand stays, a CSP can decide to deploy its own infrastructure.
To take advantage of this kind of implementation, CSP should take into consideration solutions that:
- can be implemented on any cloud
- are hardware agnostic and cloud-native by design
- is proven to run on a public cloud — but note, just because a software is cloud-native, doesn’t mean it will automatically work on any public cloud.
Innovate with SaaS and network slicing
Innovations that are the most important are those that enable any new services to run, so CSPs can profit from their investments.
Shifting the core’s consumption model from a facilities build to Software-as-a-Service (SaaS) is an innovation that drives CSP profit so they can explore new opportunities that improve the end user experience, increase monetization and simplify network operations with network software hosted on public clouds.
In addition to SaaS, network slicing is a vital ingredient. CSPs can be seen as looking not only for applications and end-customer devices but also for dynamic connectivity and configurations that are optimized for each use case. These play a key role in delivering customer experiences.
Automation makes the difference
Automation is key in optimizing expenditure on operations. Cloud-native design principles allow CSPs to operate like webscalers, so that they can easily grow and shrink the network like any webscale application. To do this, CSPs need to first approach the network as a collection of orchestrated and automated microservices instead of a collection of disjointed network elements.
As important as network slicing is, the automation of slice creation, configuration and management is even more important as it enables CSPs to rapidly target and serve new customers to create new revenue streams.
Another aspect of automation is the expansion of the DevOps concept into DelOps, which you can think of as continuous integration and continuous delivery (CI/CD) extended from a hyperscaler’s environment to telecom. The DevOps models used by many web companies are inherently single vendors and do not work in a CSP’s multi-vendor environment. With DelOps, CSPs can address this challenge and bring true multi-vendor CI/CD agility to the network.
Using APIs to overcome roadblocks to new services
Achieving the revenue gains that 5G promises and the technology investment it demands are another CSP challenge. 5G-enabled technology has opened up many opportunities for new services. However, API-based services have historically been challenging to implement when the network relies on telecom-specific hardware, but the service-based architecture and cloud-native systems are now clearing these roadblocks.
A new concept known as “Network-as-Code” (NaC) will enable developers to easily create any brand new services they want through integrated development environments, putting the new services together in a chain via simplified APIs. Network-as-Code unlocks the network for developers and fosters openness to drive innovation and profit for CSPs.
Security must start from day zero
Security is critical to CSPs, but as 5G and the packet core evolve to support new services, it opens the network to be attacked in new ways. Giving the same user experience outside your region means you’re now having to deploy in locations that are not necessarily 100% within your control. Traditionally, telecom networks had closed proprietary hardware systems, but with the cloud, CSPs need to evolve their security to the same level. 5G standards and 3GPP standards put many safeguards in place using secure communication techniques, encryption and more.
Additionally, CSPs are asking packet core vendors to make sure they create a culture of security from the ground up — focusing on an approach that includes confidentiality, integrity and security to ensure that the development environment keeps up with a rapidly evolving industry. Overall, no matter how a CSP chooses to implement its packet core, ensuring a flexible deployment, reducing 5G implementation risks through public clouds, creating a fully automated environment and using an API abstraction to speed up the development of new services should be considered at the front end of the decision-making stage to ensure smooth cost-optimized operations and the best possible return on investment.
