Next-generation Radio Access Networks (RANs) are characterized by the disaggregation of the network elements and the use of intelligent controllers to manage the increased complexity of radio technologies and network deployments.
The disaggregation of the network elements to create an Open RAN that allows the equipment from different vendors to communicate requires well-defined, standardized interfaces. Furthermore, the increased complexity of the network and the need for real-time control forces the network operators to investigate automated solutions based on Machine Learning (ML) algorithms. The standardization effort is being led by the O-RAN Alliance, an international community of mobile network operators, vendors, and research and academic institutions operating in the RAN industry. NIST, also a member of the O-RAN Alliance, is well-positioned to assist the telecom industry by developing methodologies and tools to assess the interoperability and performance of Open RAN solutions.
An Open RAN gNB is a decentralized gNB and consists of three parts, each with its own hardware/software: Radio Unit (RU); Distributed Unit (DU); Control Unit (CU). Each unit in Open-RAN takes over a section of gNB tasks. The RU has the responsibility of transmitting and receiving RF data. The DU takes the LTE/NR layer 2 responsibility and is between RU and CU. Finally, the CU takes over the higher LTE/NR layer responsibilities.
The real time monitoring and optimization of the network requires the introduction of intelligent controllers. There are two types currently described in the O-RAN framework: a near-real time (RT) controller that receives event notifications and measurements information from the RAN nodes (e.g., CU, DU, and RU) and issues command to optimize the network. A non-RT controller is also defined to decide on the overall network operating policies to deploy based on information received from the near-RT controllers and operators’ intent. In each type of controllers are custom applications that process the information and decide actions to take. Operators or third parties can develop and use those applications to fit their specific needs. Because next generation RANs are so flexible, they are controlled by thousands of parameters. Combined with the ever-increasing number of towers being deployed, it is necessary to use ML-based algorithms that can react quickly to known and unknown changes in the network.
The industry and the research community face several technical challenges associated with an Open RAN system. Those challenges include:
In addition to utilizing standard 5G New Radio technology, the open interface concept aims to boost U.S. manufacturing of cellular network hardware by providing a platform for vendor interoperability; cellular carriers will no longer be locked into a single vendor when they purchase their network equipment. This architecture also helps to alleviate a critical national security risk by increasing on-shore technology development and manufacturing.
To accelerate the development and adoption of Open RAN solutions, NIST is focusing on the following areas: