Petar Popovski, Editor in Chief of IEEE JSAC and the Guest Editors of the Special Issue
Published: 7 Nov 2022
The special issue on Open RAN is currently inviting paper submissions. The Guest Editor team features both academic researchers and industrial researchers and innovators: Tommaso Melodia (TM), Michele Polese (MP), Rittwik Jana (RJ), Raymond Knopp (RK), Melike Erol-Kantarci (MEK), Falko Dressler (FD), and Mischa Dohler (MD). We have asked the Guest Editors to provide some of their views and predictions on Open RAN and its development in the coming years. Their valuable insights into this area of growing relevance are given below.
JSAC: What do you see as the most pressuring research question in Open RAN?
TM: Open RAN has the potential to revolutionize network control by exposing KPIs and control knobs through a centralized abstraction (the “RAN Intelligent Controller”, or RIC) able to run multiple control applications in the form of xApps/rApps. How to harmonize and orchestrate this multiplicity of control applications in a coherent manner toward a common objective is a key research problem that the research community will have to deal with.
MP: How can we design data-driven solutions that leverage xApps/rApps that generalize well, adapt to new scenarios/environments, and can be actually deployed on production networks? This is the fundamental question - more follow, e.g., what kind of service models and programmable interface do you need in the RAN? How do you efficiently collect data? Etc.
RJ: There are many research challenges in all layers of the Open RAN stack: cloud and infrastructure, service management and orchestration, RAN closed loop control algorithms leveraging AI, platform security and systems integration. Perhaps the most pressing research question in my mind is to develop new near-real time RRM algorithms that allow differentiation of services by an order of magnitude on a per UE basis without sacrificing improvements or fairness that already exist at the aggregate cell level.
RK: The availability of certified OpenRAN and 3GPP-compliant platforms (hardware and/or software) allowing for direct integration of innovative work from research organizations.
MEK: The most pressing issue in O-RAN comes when we see deployments in the field, that is the integration challenge. How can CSPs ensure e2e verified and assured networks, is still an open question. In terms of research, this brings in reflections from the IT world mapped onto telco world. Intelligence, aka AI, becomes very important for such complex and large-scale systems when human-driven or algorithmic approaches cannot simply digest and correlate the information flowing from the entities.
FD: Open RAN is a mediator for many new technological solutions. Research questions include whether the current Open RAN definition is sufficiently broad to cover all requirements, how it can be extended, how to help maintain critical KPI such as resilience (including robustness and latency) and energy consumption.
MD: There are three main research challenges: 1) cybersecurity, i.e. ensure the interfaces and resulting overall system are secure given the the significantly increased attack surface; 2) conflict resolution among AI-automation apps, i.e. ensure that decisions taken by several rApps does not cause system runtime conflicts; and 3) energy-efficiency, i.e. ensure that disaggregation and cloudification does not jeopardize the important energy gains made over last years.
JSAC: What is the relationship between Open RAN and 6G, which is poised to move beyond communications?
TM: Open RAN is about disaggregation, softwarization, and programmability through coherently defined abstractions. In the evolution toward 6G, I see Open RAN as enabling programmability of the sensing functionalities (beyond communication), as well as becoming a vehicle to promote disaggregation and interoperability to a higher degree than what we are starting to see in the industry today.
MP: Open RAN represents a vision that has been developed on top of existing cellular architectures - I see it as an add-on that completes 3GPP RAN/core specifications adding more capabilities, programmability, closed-loop control. With the next cycles of 3GPP specifications (i.e., the releases that will lead to advanced 5G designs and then 6G), more elements and ideas of Open RAN will trickle down to the RAN/core specifications, toward cloud-native and programmable cellular networks.
RJ: 6G should embrace Open RAN principles holistically and pave the way for new use cases that will produce a 10x improvement compared to current 5G technology. Every component (or the full stack) in 6G should be disaggregated (perhaps even layer 1 technologies). Assembly of the best of breed technologies should be possible driven from a high level intent.
RK: Even if OpenRAN based systems are ultimately only used as a set of community-based proof-of-concept tools, they can be a catalyst for integrating more academic research organizations into the 3GPP process and diversify the intellectual property offering for 6G. This can only be beneficial for the 3GPP ecosystem if the licensing used by OpenRAN solutions is aligned to 3GPP IPR policies. Ultimately the real impact on 3GPP and 6G standardization will come from inclusion of OpenRAN interfaces and testing procedures inside or in stronger collaboration with 3GPP.
MEK: These are standardization efforts driven by different entities. The whole telco evolution is part of 3GPP defined as in 2G3G4G and 5G, and finally 6G. Open RAN is coming out from O-RAN alliance which was previously started under xRAN efforts. Open RAN comes with different set of objectives, is driven by operators and is expected to influence 6G in some ways. However, it should be in a way that we don’t end up with fragmented standardization.
FD: Open RAN has already now helped shaping research towards 6G. The software architecture defined by Open RAN has been used as a blueprint for initial 6G architectural decisions. It also helps integrating new services and concepts ranging from joint communication and sensing to the use of intelligent reflective surfaces and to next generation virtualized edge computing.
MD: 6G will be defined by the ITU, allowing any technical consortium to propose technical solutions. However, it will likely be the 3GPP that will carry out the 6G design work globally. Therefore, whilst Open RAN won’t be synonymous with 6G, it is likely that Open RAN will inspire the 3GPP to adopt design principles from Open RAN in the 6G releases 20 and onward.
JSAC: It is 2030 and what happens with Open RAN? Is it used widely everywhere? If not, what could have prevented that?
TM: I believe it will be used everywhere. The Open RAN revolution will change cellular network design, deployment, and control drastically, and will accelerate innovation in this space. I believe we will start to see significant deployments in the next couple of years, starting from private 5G and moving quickly to public networks.
MP: Networks that are based on software will evolve at a faster pace than traditional cellular generations. I see O-RAN solutions being deployed much earlier than 2030, especially in some markets like private 5G.
RJ: Looking through the 2030 crystal ball, Open RAN procedures will get more sophisticated. Disaggregation will be the bedrock principle for all facets of mobile communications (Open RAN, Open Transport, Open Core, Open UE). Complex services can be automatically composed, deployed, activated and monitored using any “Open” technology in a matter of minutes. The overall ecosystem will be fully self healing with little human in the loop intervention.
RK: There is no doubt that OpenRAN will help diversify the greater 3GPP ecosystem. It is already happening today with the availability of O-RAN O-RU for smaller players simplifying the generation of impact in an extremely competitive arena comprising primarily major industrial actors. That in itself is a success. Massive deployment of OpenRAN hardware and software components for the networks of mainstream operators will depend largely on the relationship between the main OpenRAN players and the 3GPP through the alignment of roadmaps in the 6G timeframe. Vertical industries betting on private xG networks will likely be able to benefit from OpenRAN solutions, because of the ability to tailor software to their specific needs. This may end up as a key commercial success for OpenRAN if private xG itself is successful.
MEK: The telco space is evolving much faster than ever. It’s harder to look at a crystal ball and predict it’s future. The facts are 3GPP plans for 6G is on the way and we expect deployments of 6G in the field at that time. Open RAN’s horizon is pretty much reliant on the abilities around integration, security, performance and energy-efficiency. With 5G and the continuing generations, the wireless has become a critical infrastructure and one cannot drop the ball in any of the above criteria.
FD: Open RAN is currently helping to define fundamental 6G architectural concepts. So, Open RAN concepts will become integral part of next generation networks. Whether these concepts survive under the brand Open RAN or independently as part of 6G and beyond standards, does not really matter.
MD: The commercially viable principles of Open RAN will likely be widely adopted, such as cloudification, (most) open interfaces and decoupled intelligence via rApps/SMO. It is also very likely that these leading principles will be integrated into 3GPP releases.
Statements and opinions given in a work published by the IEEE or the IEEE Communications Society are the expressions of the author(s). Responsibility for the content of published articles rests upon the authors(s), not IEEE nor the IEEE Communications Society.