Patrik Rokyta I CTO I January 21th, 2025
Over the past six years, most communication service providers (CSPs) and telecom suppliers have embraced cloud-native practices. These practices, initially adopted for 5th Generation (5G) mobile core networks, are now becoming the preferred infrastructure and deployment model across various telecommunication systems. This shift is fundamentally transforming how services are delivered, managed, and experienced.
In this blog, we explore key terms and trends in deploying telecom products and solutions, addressing both infrastructure-related aspects and the experiences with different deployment models.
From Network Function Virtualization (NFV) to Cloud-Native Network Functions (CNFs)
The transition to cloud-native began with the industry standards for the 5G mobile core network, moving from virtual network functions (VNFs) to cloud-native network functions (CNFs). CNFs consist of microservices—small, modular software components running as containers on a host operating system.
Microservices are designed for single-purpose functionality, with each component using isolated resources (CPU, memory, storage). This isolation reduces the risk of error propagation and resource exhaustion across the system, enhancing reliability. Moreover, processes such as continuous integration, delivery, auto-scaling, and in-service upgrades significantly improve operational efficiency, making these processes appealing irrespective of the mobile core version. With this, CNFs are showing up across various subsystems and domains such as 4G, 5G, and IMS—including number-portability corrected ENUM, infrastructure and public DNS.
Compared to VNFs, deploying CNFs introduces additional complexities. Managing the individual microservices of multiple CNF instances requires container orchestration frameworks like Kubernetes. Kubernetes adds a management layer to the deployment stack, improving resilience and security through features such as elasticity, redundancy, self-healing, network policies, canary rollouts, and graceful shutdowns. Additional cloud-native services must be deployed alongside Kubernetes. Examples include but are not limited to container image registry, cloud-native storage, networking and service exposure. Yet, despite this complexity, CNFs clearly represent the foundation for resilient, scalable, and future-proof telecommunication ecosystems.
The Evolution of Telecom Infrastructure
Traditionally, on-premises computing has been and is the dominant approach for implementing mobile networks. In addition to data center computing enabling efficient management of centralized services like authentication and subscriber data management, edge computing has emerged to support services that require proximity to data sources for reduced latency and localized processing.
The advent of cloud computing introduced a transformative paradigm. Cloud technology offers flexibility, on-demand scalability, automation, orchestration, rapid innovation, and continuous delivery. These advantages have gained significant traction, just recently in AI/ML applications.
To bridge the gap between on-premises reliability and cloud-native efficiency, platforms like Red Hat OpenShift, VMware Tanzu, and the open-source Sylva project enable telco cloud experiences while maintaining physical control over infrastructure. Recognizing this demand, hyperscalers such as Amazon, Google and Microsoft have also entered this market with private and hybrid cloud offerings, aiming to deliver similar capabilities while integrating seamlessly with public cloud ecosystems.
Deployment Challenges and Opportunities
The rule of thumb is straightforward: VNFs should be deployed in virtualized environments, while CNFs are best suited for cloud-native ecosystems. However, a significant challenge remains: managing and integrating legacy systems. For CSPs, the challenge lies in adopting a new cloud-native infrastructure while simultaneously maintaining their existing virtualized environments. For telecom suppliers, the difficulty is in transforming legacy VNFs into CNFs, particularly as CSPs aim to standardize their infrastructure and workforce around a unified, future-ready technology: cloud-native solutions. These transformations are progressing unevenly across the industry and between operators and suppliers, leading to heterogeneous deployment models. Examples include running containers within a virtual machine, deploying legacy virtual machines (VMs) within cloud-native ecosystems (mainly using a "lift and shift" approach), or even porting selected (primarily 5G) CNFs back to VNFs.
Mobile network operators and IPX service providers that combine telecom operational excellence with the agility of technology companies are often called “techcos.” Techcos have embraced cloud-native technology, building infrastructure capable of running and orchestrating containerized workloads (CNFs). Many of these infrastructures also natively support VNFs, alleviating immediate pressure on incumbent suppliers to transform their legacy product portfolios into CNFs. In the mid-term, suppliers are expected to decompose their VNFs into microservices-based CNFs, adopt cloud-native best practices, and implement processes tailored for container-based workloads. Conformance specifications and tests exist to validate whether a CNF adheres to cloud-native principles.
Other operators may initially focus on addressing challenges introduced by 5G, such as managing the new signaling protocol, negotiating 5G SA roaming contracts, or testing 5G PLMN interconnects. These operators may prefer utilizing their existing virtualized infrastructure even for 5G, especially if they do not have an immediate need for elasticity or auto-scaling. On the supplier side, the capability to deploy 5G CNFs in virtualized environments is often achieved by deploying virtual machines with containers running in an unmanaged way, thereby forfeiting the benefits provided by container orchestration platforms like Kubernetes. Notably, many CNF suppliers have already tightly integrated their CNFs with Kubernetes, leveraging patterns like Kubernetes operators to enhance provisioning, configuration, and lifecycle management. These CNFs have become dependent on Kubernetes and can no longer function without it. Consequently, Kubernetes must be delivered and maintained by the supplier alongside such CNFs. Additionally, all cloud-native aspects—such as container image registries, cloud-native storage, and service exposure through load balancers—must also be managed entirely by the supplier.
The Road Ahead: Trends for 2025
By 2025, various deployment streams are expected across the industry.
For mobile network operators and IP interconnect service providers:
- Bare-Metal Containers: Deploying containers directly on bare metal eliminates the virtualization layer, delivering superior performance, resource efficiency, and a simplified architecture.
- Heterogenous Deployments: Operators will aim to harmonize their infrastructure, with cloud-native technology becoming the preferred approach. Third-party solutions, such as Red Hat OpenShift Virtualization, will support running VNFs within cloud-native environments, ensuring flexibility during the transition.
- Hybrid Cloud Adoption: Hyperscalers will continue to attract telecom operators to the cloud. Hybrid clouds, which combine selected services deployed on public clouds with others retained on-premises, will emerge as a practical and balanced solution. Resource-intensive applications such as AI/ML, optimized for cloud environments, will drive these deployments, with a strong focus on data security and compliance.
For suppliers:
- Advanced CNFs: Suppliers will enhance the operational experience by implementing Kubernetes patterns like operators delivering improved lifecycle management, and by exposing services through standardized APIs, encouraging value-added services development.
- Continued VNF to CNF Migration: Suppliers delivering products for 4G, 5G, IMS core networks and even beyond will be expected to accelerate the migration of their legacy portfolios to CNF architectures. This migration will require decomposing monolithic VNFs into microservices-based CNFs, fully aligned with cloud-native best practices, rather than relying on a simple “lift and shift” strategy. Depending on market demand and business opportunities, certain CNFs may be designed from the ground up to fully leverage cloud-native technology and its associated processes.
About Titan.ium
Titan.ium Platform is a leader in signaling, routing, subscriber data management, and security software and services. Our solutions are deployed in more than 80 countries by over 180 companies, including eight of the world’s top ten communications service providers.
Titan.ium began its cloud-native journey in 2019 with the introduction of its Titan.ium cloud-native platform. At the same time, we continue supporting the Titan virtualized platform that can also be deployed on physical servers.
By the end of 2024, Titan.ium’s cloud-native portfolio includes several 5G network functions (SEPP, NRF, SCP, NSSF and BSF) and selected legacy network functions (SS7 and Diameter signaling firewall, DNS, ENUM and Ut-Proxy) that have transitioned to cloud-native to address immediate market demands. The transition process is ongoing for the remaining legacy products in the Titan.ium’s product portfolio, with Diameter Signaling Controller (DSC) being the recent addition and Number Portability (NP) server next in line. This gradual shift enables CSPs to harmonize their infrastructure while ensuring continuity.
