Summary Bullets:
- The Open Networking Foundation (ONF) has launched a ‘Software-Defined Radio Access Network’ project aimed at developing open-source RAN solutions using an ‘app store’ model for network optimization features.
- This effort will be helped by the open RAN (ORAN) and virtual RAN (vRAN) movements now gaining steam, but it will also confront some of the same hurdles facing open RAN – including opposition from incumbent major vendors.
The Open Networking Foundation – a group that promotes open-source networking technologies – announced a ‘Software-Defined Radio Access Network’ project aimed at developing open-source RAN solutions.
The project is focused initially on the creation of a near-real time Radio Intelligent Controller (RIC) that will apply an app store model to RAN functions. The ONF envisions an ecosystem of competing software vendors making RAN control and optimization applications available for the open-source RIC. The foundation is kicking things off by offering its own apps governing subscriber handover (how the network passes a user from one cell to another) and load balancing (overburdened cells transferring traffic to less congested cells). It has already demonstrated its own prototype’s control of 100 emulated base stations and hopes for field trials by early next year.
The ONF has a solid track record of bringing open-source architectures to other parts of the network (optical transport, packet switching, etc.), but it was late to the RAN for many of the same reasons that network virtualization was slow to arrive at the RAN: the unavoidable importance of hardware at the network’s end. However, the advancement of the open RAN movement (disaggregating the RAN into centralized and distributed baseband units as well as radio units that can be sourced from different vendors) helps free the RAN from its hardware shackles, fueling RAN virtualization and, in turn, making it easier for the ONF’s RIC – which supports open RAN interfaces – to play a role.
It’s not surprising, then, that the arguments ONF is making for the value of its RIC closely mirror the arguments for open RAN and vRAN. As geopolitical trends have erected barriers to Huawei and ZTE in some markets, mobile operators have been left with fewer choices for the RAN – one of their biggest network expenditures. Open RAN was proposed by major operators around the globe in order to encourage greater competition among RAN vendors, and so far, it looks like it’s working: new entrants have joined the market on both the radio hardware side (e.g., BaiCells, MTI, NEC, etc.) and, thanks to vRAN, the software side (e.g., Altiostar, Mavenir).
Many of the same major operators that support the open RAN movement (AT&T, China Mobile, etc.) are also supporting the ONF’s SD RAN project, imagining a vibrant marketplace of RAN control app providers to drive innovation, network efficiency, and cost reduction – and to help enable new use cases in the enterprise domain. In 2020, we’ve seen vendors such as Nokia, Samsung, and others responding to operators’ push for open RAN; the fact that ONF is working with Radisys, a widely influential RAN supplier, is another example of the wind in open RAN’s sails.
The ONF should expect to encounter some of the same hurdles as the open RAN movement, too. First, it should count on opposition from the entrenched major RAN vendors such as Ericsson and Huawei, since a RAN function app store represents a competitive threat to their traditional business. Although the ONF can proceed without those vendors, it won’t be easy, since they wield significant power – in their operator and component-maker relationships, their patent portfolios, and their influence over industry technology standards.
Second, some operators will prefer the simplicity of turnkey solutions from trusted suppliers over architectures that require them (or an integrator) to smoothly stitch together a patchwork of disparate network elements and functions and manage a larger array of supplier relationships. Operators will likely have concerns about whether such networks will be reliable, efficient, and secure – and whether their added complexity will add operational cost. To its credit, the ONF can say that it has heard these concerns before – when it brought open source to optical transport and packet switching – and proved skeptics wrong.