Is the O-RAN concept inevitable in the network evolution path?

I have had many discussions with my colleagues from telcos about the O-RAN RAN ( Open Radio Access Network) implementation perspective on their networks. All have one thing in common that the O-RAN concept doesn't have the chances due to the immaturity, performance limitations, etc.

Need to mention that their viewpoint is narrowed down to the choice of swapping the existing RAN with the O-RAN .The O-RAN movement has a broader effect on the RAN market.

?Let’s conduct a high-level analysis to determine whether my colleagues' pessimism or the forces of development are stronger.

Such pessimism often arises from the need to balance innovation with reliability and the proven effectiveness of existing technologies.

??There is an undisputed trend in telecom towards flexibility and customized services. This became inevitable when data services took a leading role over voice services, significantly challenging the commonly adopted 'one-size-fits-all' approach to service provision and optimization

The?conceptual solution was to move all service provisioning on software level. All the development in industry has being actively moved in this direction for last decade.

With the goal to accelerate transformation, TMForum developed a long-term vision and reference architecture in the Autonomous Network project that went public in 2017. Different domains should develop hand in hand to achieve this breakthrough.

?Let’s examine these trends in the context of network evolution and assess how O-RAN aligns with them.

?

What we have on Core Network side?

??4G was the first mainstream mobile network that was designed natively for data service benefiting from?more flat architecture.?The next milestone was the EPC virtualization (vEPC) on the ETSI NFV ISG Architectural Framework. Finally, the Control and User Plane Separation (CUPS) of EPC was introduced in Release 14, marking a significant milestone towards further flexibility.

The core network architecture was inevitably moving towards the slicing concept, which involves multiple logical networks tailored to specific sets of requirements or use cases.

?Reference to experience on attempts to get the slicing benefit

I have delivered numerous projects aimed at overcoming flexibility limitations in pre-5G networks. These projects included solutions that utilized Network Function Virtualization (NFV) by provisioning additional vEPC for creating isolated solutions, particularly for M2M services. We implemented Traffic Engineering solutions to leverage QoS and DPI options. Our efforts also focused on overcoming the aggregation level parameter limitations that applied to all services. Most of it achieved the goals at some level due to lacks of alternatives

?5G Core network (5GC) is based on Service-Based Architecture (SBA) and?has native slicing solution?in contrast to EPC. 5GC has also additional option leading to?paradigm shift.?5GC is designed with Network Exposure Function (NEF) that opens?the network capability to third-party applications?so we should perceive 5GC as platform not as Core Network in traditional viewpoint.

I can?predict the emergence of new ecosystem around 5GC (APP Store for 5GC) to leverage the?NEF and Software Development Kits (SDK)

The introduction of vEPC and 5GC shifted the competition to the software (SW) area, leading to the emergence of new services such as AWS Telco Network Builder (AWS TNB), Google Cloud for Telecom, IBM Telco Network Cloud Manager, and projects like the Magma Project under the Linux Foundation. These initiatives aim to accelerate the adoption of private networks, which is one of the promising use cases for 5G today.

?

In summary, the affordability of 5GC is opening doors for new players across various verticals. I foresee a shift from a nationwide 5G network deployment approach to a more fragmented one.

Assuming above mentioned, the?demand for diverse?? RAN? software-defined solution?is huge. The question arises: Can this be achieved with proprietary solutions, or is an O-RAN with an open architecture implementation necessary?

?What we have on RAN side?

?RAN solution?under huge pressure to be more flexible and cost effective.?

As I previously mentioned, full flexibility is realized in a software-defined world where open APIs enable contributions from a diverse developers

?In telecommunication term it means that?RAN needs to be both?virtualized and?programmable.?This journey began with Software-defined Radio (SDR) introduction.

?I experienced this flexibility personally?when we swapped legacy RAN on new SDR RAN and the launch of NB-IoT and LTE were the pure configuration scope. It was matter of weeks. I really understood the full power of Software-defined solutions

?The latest?RAN domain concept is Cloud RAN?(C-RAN) that is a viable option??to increase flexibility, speed up the delivery of services and enable the greater scalability.

C-RAN represents a form of RAN virtualization. In this model, the baseband RAN functionality is delivered entirely as software, decomposed into a modular architecture.

What about openness ?

C-RAN, however, is not designed with open APIs, which restricts the capability of third-party developers to introduce completely new approaches to existing problems.

?C-RAN can be integrated on the current stage with service orchestration platforms?such as ONAP (Open Network Automation Platform) or ETSI MANO?through standard APIs.?But it is all about automation and efficiency leveraging existing proprietary functionalities.

?What is potentially new value creation is natural synergy of C-RAN and Mobile Edge-Computing?(MEC) where C-RAN?provides a platform for hosting MEC APP and?other VNFs.

?

What O-RAN brings on the table?

?The most transformative benefit of Open RAN is introduction of?two flavors of RAN Intelligent Controller (Non-RT RIC and Near-RT RIC) to get the RAN programmable.

?RIC brings the declarative?Intent-based services management and paves the way for a?larger ecosystem of?rApps developers

This will significantly enhance automation and offer opportunities to tailor the network services lifecycle to specific local contexts. Given the increasing complexity of RAN, both automation and customization are essential to fully benefit from emerging technologies like DSS, Massive MIMO, Network slicing and 256QAM

?When I led the optimization project for the IIHF World Championship, the biggest challenge was managing the capacity of hot spot areas. This was particularly difficult because admission/overload control and traffic layering features had to be manually adjusted on a timescale of hours. Consequently, my team was always several steps behind

I wish I had access to intent-based configuration with pre-tailored rApps, AI-based optimization loops, and massive MIMO technology, enabling me to adjust a cluster of sites on a timescale of seconds

?On one hand, O-RAN brings flexibility as one of its differentiators; on the other hand, there are no clear answers to the question of how that flexibility can be monetized.

Other advantages like?vendors unlock and possibility to get the huge ecosystem of specialized vendors on RAN side my colleagues see as?risks now and just as ?the long-term vision.

I believe there are no strong incentives to adopt commercial O-RAN solutions at scale in the next few years, given the significant shift they would bring to the deployment paradigm.

I foresee that telecom companies will continue their global 5G deployment using well-known partners with C-RAN solutions. These vendors will deliver their own RIC platforms and offer their own SDKs, which may not necessarily interoperate with each other.

O-RAN pushes C-RAN to become open and it is inevitable effect. It is not SWAP kind of desicion

Conclusion

In conclusion, it appears that my 'pessimistic' colleagues are right to some extent. But, O-RAN is also driving C-RAN towards opennes, and this shift seems inevitable. It's not a 'swap' in technology decision, but a fundamental change.

However, O-RAN has the potential to challenge the current market. For example, European operators such as Deutsche Telekom, Orange, Telefónica, TIM, and Vodafone have published further documents on technical priorities for O-RAN in 2023. Many countries view O-RAN as an opportunity to develop domestic 5G networks.

What is your opinion? Please share it in the comments