Open RAN is changing how radio access networks are designed by separating functions that were traditionally delivered as tightly integrated systems. Radio units, distributed units and centralised units can increasingly be deployed as separate components, creating greater flexibility in how network functions are built and managed. This architectural shift is also increasing the importance of cloud infrastructure, as software-based RAN functions require computing environments capable of supporting distributed workloads.
This is where cloud native Open RAN becomes relevant. Rather than tying network functions permanently to specialised hardware, cloud-native architectures allow software to operate across commercial computing infrastructure. The approach can support more flexible deployment, software-based upgrades and the sharing of computing resources across different workloads. It also brings practices developed in the broader cloud industry, including containerisation, orchestration and automated lifecycle management, into the RAN environment.
The market opportunity is becoming more visible, although the transition is not happening at the same pace across all networks. Dell’Oro projected that Cloud RAN could account for nearly 25% of the total RAN market by 2029. At the same time, the research firm has continued to highlight performance, power consumption and cost-parity challenges as factors influencing adoption. This suggests that moving RAN functions onto cloud infrastructure is not simply a matter of replacing dedicated hardware with commercial servers. The underlying infrastructure must still meet the demanding requirements of carrier-grade networks.
Virtualisation is Changing the RAN Infrastructure
In conventional RAN environments, processing functions are closely associated with purpose-built equipment. Virtualisation separates software from that dedicated hardware, allowing network functions to operate on standardised computing platforms. Cloud-native architectures take this further by treating those functions as software workloads that can be deployed, managed and updated through common cloud-native tools.
Kubernetes is becoming an important part of this shift. The Cloud Native Computing Foundation identifies Kubernetes as a key platform for cloud-native network functions, while the O-RAN software ecosystem is increasingly using Kubernetes-based tools to deploy and manage components of the RAN architecture. This brings capabilities such as automated deployment, workload scheduling, scaling and recovery into an environment that has traditionally relied more heavily on specialised network hardware.
The change also affects infrastructure planning. Computing capacity can potentially be distributed between locations depending on the requirements of each RAN function. Some workloads may need to remain close to the network edge because of latency or processing requirements, while others can be handled in more centralised facilities. This creates greater flexibility, but it also makes infrastructure placement an important engineering decision.
Deployment Models are Becoming More Flexible
Cloud-native infrastructure allows operators to consider different ways of distributing RAN workloads across the network. Processing can be positioned closer to radio sites, within regional facilities or in larger centralised data-centre environments depending on traffic patterns, latency requirements and available computing resources.
This flexibility is particularly relevant as operators manage networks with very different levels of traffic and capacity demand. A heavily loaded area may require additional computing resources, while a lower-demand location may not need the same level of dedicated infrastructure. Software-based deployment can potentially make these resources more adaptable than fixed hardware architectures.
The O-RAN architecture is also developing around the concept of the O-Cloud, which provides the infrastructure environment for cloud-based RAN functions. O-RAN specifications include the O2 interface for cloud infrastructure management and workload orchestration, reinforcing the role of cloud infrastructure as a distinct layer within the RAN architecture.
This changes the strategic role of computing infrastructure. It is no longer simply the hardware underneath the network. It becomes part of the mechanism through which RAN functions are deployed, scaled and managed.
Cloud Native Open RAN therefore represents more than the virtualisation of existing network functions. It points toward a deployment model in which software, compute and network resources can be managed with greater flexibility. The potential benefits include faster deployment, more adaptable resource allocation and greater separation between network software and hardware. However, the ability to deliver those benefits at carrier scale will depend on whether cloud-native infrastructure can meet the performance, energy and cost requirements of modern mobile networks.
Key Takeaway: Cloud RAN could represent a meaningful share of the RAN market by 2029, but performance, power and cost considerations are likely to influence how quickly cloud-native deployment expands.
The result is a more complex deployment decision for operators. Cloud native Open RAN can create greater flexibility in how network workloads are distributed, but that flexibility has to be balanced against latency, energy efficiency, processing performance and infrastructure cost. As these trade-offs are better understood, cloud native Open RAN is likely to develop through a mix of deployment models rather than a single universal architecture.
The broader shift is therefore not simply from hardware to software. Cloud Native Open RAN is creating a new infrastructure layer in which compute resources, network functions and orchestration systems increasingly need to work together as a single operational environment.