加速部署的开放RAN DU测试注意事项

Open RAN DU TestingConsiderations forAccelerated Deployments Open RAN DU TestingConsiderations forAccelerated Deployments Executive Summary The 5G network is becoming more automated, distributed, andsoftware-driven, presenting new opportunities. Virtualizedstandalone networks that enhance service offerings whileoptimizing Opex and Capex are of increasing interest tooperators. In particular, communications service providers wantto understand how Open RAN architectures that support multi-vendor interoperability and standardization can enhance overallnetwork capabilities. This also applies to Open virtual RAN (OpenvRAN), with the principles and benefits covered in this white paperrelevant to both Open RAN and Open vRAN. Operators understand successful transformation is only possibleif they can confidently deploy Open RAN’s disaggregatedcomponents (i.e., the CU, DU, RU, and RIC). As early trials anddeployments have shown, each presents challenges and requiresrobust testing to ensure seamless functionality. The adoption ofCI/CD/CT practices introduces further challenges, demandingdynamic testing solutions that can support rapid deploymentcycles. Sophisticated testing mechanisms are required to effectivelynavigate Open RAN deployment complexity and support operatoreagerness to diversify network architectures in preparation for anew era of mobile services delivery. In this paper, we focus specifically on strategies for testing andvalidating the O-DU given its important role in supporting real-time network functions for high-quality service delivery. We closeby examining Spirent’s methodology for comprehensive andautomated testing that addresses Open RAN and the O-DU’sunique challenges. RAN systems. Network functions run as virtualizedsoftware on open cloud infrastructure that reducesOpex through extensive automation. The RAN’sdisaggregation and virtualization allow flexibledeployments, deliver new use cases, and stimulateincreased innovation. Accordingly, the term OpenRAN in this white paper encompasses both OpenRAN and Open vRAN architectures. The Disaggregation andCloudification of the RAN The 5G mobile core network is transforming intoa distributed, disaggregated, software-driven,cloud-native architecture. These 5G virtualizedstandalone networks increase revenues with newservice offerings. They also improve efficiency andreduce Opex and Capex by increasing agility andautomation, and using commodity hardware. Open RAN’s layered software stack is agile enoughto support a range of specific service and operatorneeds, as well as fast development, testing, andupdating (DevOps). Embedded intelligence in everylayer enables dynamic resource allocation andoptimization. Open RAN is ushering in a similar revolution tothe radio access network. By disaggregating andvirtualizing the RAN, mobile network operatorsexpect to reduce Opex and Capex. Open RAN’s openarchitecture creates a standardized frameworkas defined by the O-RAN Alliance with interfacesdesigned to diversify the monolithic RAN supplychain and support multivendor interoperability. The Open RAN architecture disaggregates thetraditional monolithic RAN base station into thecentralized unit (CU), distributed unit (DU), radiounit (RU), and RAN intelligence controller (RIC), withmany new interfaces. These open and interoperable architectures arepowered by modular, layered software runningon off-the-shelf hardware for dynamic scalabilityand reliability at lower costs versus traditional WHITE PAPER Open RAN cloudification The mobile industry’s journey to cloud-native began with monolithic systems and applications that were tightlyintegrated with vendor-specific hardware. As mobile infrastructures have become software-based, they havegained the flexibility to run on commodity hardware. Over time, these functional building blocks, known as network functions, have progressed from physical to virtualand now, cloud-native, with key distinctions at each evolutionary step: •Physical.Single vendors provide network nodes or functions (e.g., a base station) with integrated hardwareand software.•Virtualized.Virtualized network functions are disaggregated from underlying virtual machines to improvehardware utilization but rigid coupling between network functions and virtual machines limits flexibility.•Cloud.Applications reside in the cloud and are accessed by virtualized network functions.•Cloud-native.Containerized cloud-native functions communicate via standardized APIs, with microservicesabstracted and exposed to smaller units of code (pods) that are orchestrated with Kubernetes. A cloud-native architecture enables automation and increases network efficiency and hardwareutilization by: •Breaking software into smaller piecesin alignment with microservice principles so more network functionscan run on existing hardware and serve more customers with the same hardware investment.•Automatically adding or removing podsto efficiently and quickly meet demand changes.•