A massive increase in multimedia consumption, a growth in digital servi­ces, higher uptake of user engageme­nt platforms, and increasingly complex and dynamic traffic patterns are imposing new requirements on modern networks. Operators and vendors around the world are trialing and deploying open radio acc­ess networks (RAN) to modernise their net­­work infrastructure. The advantages of this are manyfold, from a reduction in the total cost of ownership to opening up of new avenues of innovation and a diversifi­ed supplier ecosystem.

Although open RAN is yet to achieve mass commercialisation, global adoption is expected to bolster in light of the recent technological developments in the doma­in. Operators are leveraging next-generation technologies such as artificial intelligence (AI), machine learning (ML), virtualisation and data analytics to accelerate the auto­mation of software components of the op­en RAN architecture. The integration of these innovations with open RAN creates numerous benefits. These include enhan­ced operational efficiency, more intelligent and scalable networks, accelerated time-to-market of new services and functions, and CAPEX and OPEX efficiencies.

A look at how these technologies are augmenting the open RAN landscape…

AI/ML

AI and ML are key to intelligent management and operation of networks for add­ressing the increasing network complexity. These technologies are being leve­raged by mobile network operators (MNOs) worldwide to significantly im­p­ro­ve the network performance of open RAN and automate operations. These te­ch­nologies consist of algorithms that learn from experience and continuously impro­ve the system. The al­go­­rithms further enable the operators to analyse enormous amounts of real-time tra­­ffic data and network load, without compromising network capacity.  Additio­nally, these technologies enhance the end-user experience by enabling telecom servi­ce providers (TSPs) to provide the quality of experience demanded by consumers and enterprises. Thus, allowing TSPs to adj­ust network conditions and ensure pro­per load balancing, and, in turn, offering seamless transfer of active call or data sessions bet­ween channels or networks. Some use cases of these technologies in the open RAN do­main include forecasting parameters, de­tection of anomalies in the system, traffic steering and prediction of failures.

Further, the O-RAN Alliance introdu­ced the RAN Intelligent Controller (RIC), which is an entity of open RAN architecture responsible for RAN general operations as well as constant optimisation. It pro­vides advanced control functionality wi­th the support of data-driven approaches such as AI/ML tools to improve resour­ce management capabilities. Leading op­en RAN solutions offer natively built-in AL/ML capabilities into the RAN architecture for the non-real time (non-RT) and near-RT RIC functions. ML models are trained in non-RT RIC rApps based on data received from different functions in the network. These are then used for analysis and to make inferences so as to provide the best end-user experience. Wi­th the commercial roll-out of 5G networks, AI and ML will have a much more comprehensive and transformative impact on the end-to-end network in an open RAN environment.

Zero-touch provisioning

Zero-touch provisioning performs automated software installation and configurations, eliminating the need for manual in­ter­vention. Besides significantly reducing complexity in provisioning, it minimises ti­me, errors and cost. Zero-touch provisi­oning is considered safe for RAN installation, and significantly helps in the deployment of hundreds of sites without requiring a site visit. While it will take time to drive automation towards zero-touch, it will be critical for dense deployment of 5G networks when a large number of sites will need to be configured.

CI/CD

Continuous integration and continuous delivery (CI/CD) is a methodology desig­ned to accelerate the software development process and delivery with more frequent product releases. The presence of different components from various vendors makes CI/CD essential in the open RAN space. Through automation, CI/CD software all­o­ws upgrade/downgrade in seconds or minutes with minimal human intervention, the­reby substantially reducing cost, time and errors as compared to site visits for testing and upgrades. Further, these updates can be monitored to evaluate their impact on end users and whether they are helping in achieving business goals.

The CI/CD approach has been recognised as a critical element for the software requirements of the RAN ecosystem. Lea­ding cloud-native open RAN solutions pr­o­vide CI/CD frameworks to enhance agi­lity and flexibility for MNOs to launch new features, fix bugs in the system and im­ple­ment software upgrades. Impleme­n­ta­tion of CI/CD software in the open RAN architecture can enable MNOs to co­llaborate with different ecosystem participants, thus fostering innovation.

Edge computing

Continuous network deployment processes such as CI/CD will drive the demand for edge computing for faster execution fr­om network edges. Key benefits of deploying open RAN systems at the edge include lower latency, an enhanced user experience, alleviated network congestion, optimised network efficiency and locally enabled computing power for devices. Through these benefits, it opens up av­en­u­es for dense 5G use cases such as au­tonomous vehicles and internet of things (IoT), and enables immersive experiences. Further, disaggregated network in open RAN allows computing power to be pushed to the edge, while other processes remain at the core. Telcos can thus design and provide the network based on end-user applications.

Virtualisation

Virtualised RAN (vRAN) is an agile approach to the deployment and management of RAN that virtualises RAN network functions and deploys them on cloud platforms. It employs network functions virtualisation that helps operators to control and route cellular resources more efficiently, which is crucial for 5G core networks. This allows interoperability among RAN components and enhances supply chain security. Moreover, virtualisation offers improved operational efficiency by using industry-established components for common tasks, thus eliminating the need for vendor-specific solutions and allowing operators to focus on business-specific components. It also provides greater flexibility in terms of functionality and capacity to meet the demands of 5G use cases. Finally, a wide adoption of vRAN can address the barriers in cross-domain innovation, leading to the development of new use cases and services.

According to a report by Analysys Mason, vRAN is the fastest-growing segment of cloud investment for operators, with spending forecast to reach $11 billion by 2025 at a compound annual growth rate (CAGR) of 132 per cent. Meanwhile, a study by Mavenir reveals that 95 per cent of MNOs understand the importance of virtualisation and have it on their open RAN roadmaps.

Cloud

A cloud-native open RAN solution en­ab­les automation by taking over infrastructure management and operations, handling the management of computing, storage and network resources. Cloudi­fi­cation of RAN essentially means the virtualisation of RAN services in containers and mi­cro services to ensure quick decision-making at the edge of where the RAN equipment is set up. Each micro service can be dep­loyed, upgraded, scaled and restarted independently of other micro services in the RAN application, thus allowing frequent up­dates to live applications without im­pacting service level agreements. By en­abling automation, the integration of clo­ud in the RAN architecture offers nu­merous benefits such as a reduction in ad­mi­nistrative overhead costs, greater re­so­urce utilisation, scalability and efficient po­wer consumption. Disaggregating RAN software and hardware in the cloud opens up the possibility of independent innovation in hardware and software.

Cloud RAN will soon adopt leading practices and become a foundation for op­enness and innovation in the 5G era. The involvement of big public cloud pro­vi­ders, or hyperscalers, such as Amazon, Google and Microsoft, will be a significant driver of the open RAN initiative.

Data analytics

Data analytics is deployed as rApps in the non-RT RIC and uses big data to provide an overview of the network conditions. An­a­lytics provide a visual representation of patterns or abnormalities in the netwo­rk, thus enabling MNOs to comprehend and draw inferences on fixing and improving network performance for a better user experience. It facilitates the review of AI data and the generation of reports on ML performance in network improvement. To optimally leverage data analytics, big data and data mining in the open RAN ecosystem, better openness and application programming interfaces between vendors and MNOs will be required.

Summing up

Despite certain challenges such as high integration complexity due to multi-vendor environments, security concerns, and integration issues for brownfield operato­rs, open RAN-based solutions are stea­dily gaining momentum. As 5G and IoT use ca­ses continue to mature, RIC and au­to­mation will become increasingly essential in the open RAN environment. The integration of new-age technologies in the RAN framework will enable MNOs to expand their networks into new areas and fulfill the promise of bringing 5G within the reach of underserved markets. The op­en standards promoted by the O-RAN Alliance will further help leverage technologies such as AI/ML and real-time analytics to develop open interfaces and faster de­ployment of open RAN. The trend to­ward these technologies is set to create new dynamics and opportunities in the open RAN ecosystem.