By Bronwyn Howell
In the mobile telecommunications industry, well-known for buzzwords and acronyms, the hottest new phrase is “open radio access networks” (O-RAN). O-RAN is an industry concept proposing an alternative way of designing and building radio access network (RAN) portions of mobile networks — both visible parts on towers and less visible parts in base stations that digitize and transmit signals between mobile phones and the network core. Whereas traditionally, network operators source RAN equipment as an integrated package from vendors like Nokia, Ericsson, Samsung, or Huawei, O-RAN envisages the disaggregation of RAN componentry into constituent parts capable of interconnecting seamlessly using open protocols and standards.
The O-RAN concept claims many advantages over historic procurement models. First, O-RAN will supposedly lead to cheaper network deployment, as multiple vendors can compete to supply individual components and network operators can mix and match on price and quality. Similarly, innovations in small component elements can allegedly be implemented quickly rather than waiting for a more comprehensive rebuild. Analysys Mason purports these advantages would “generate USD285 billion in GDP impact globally by 2030, and over USD91 billion annually thereafter.”
Second, O-RAN is closely linked with national edicts to remove Huawei equipment from mobile infrastructures. Huawei is an advocate for O-RAN as the optimal structure going forward, but as protocols and standards for operational deployment take time to develop, some allege this approach will prolong Huawei’s involvement in countries seeking to ban it. This has also led non-Chinese stakeholders to advocate for government subsidies to fund O-RAN research and development (R&D) to expedite replacement of Huawei equipment. These calls have also been associated with lobbying for governments to mandate O-RAN for 5G deployment, along with further subsidies and favoritism for local component-making firms. This has succeeded somewhat in the US with the creation of the Public Wireless Supply Chain Innovation Fund, which gives grants to “promote and deploy technology that will enhance competitiveness in 5G supply chains that use open and interoperable-interface RANs.”
Third, O-RAN has been claimed to offer improved security over single-vendor systems with “more modular, more visible and . . . less interdependent” networks. Still, it is unclear how the complexity of mixing and matching will secure a network against potential “rogue” elements rather than having one manufacturer responsible for this element of system design.
Are these claimed advantages credible? For policymakers facing lobbying from O-RAN advocates, some skepticism is warranted.
First, O-RAN is not a new concept. For years, “single-system” vendors such as Nokia and Ericsson acted as agents of network operators by bundling components designed and manufactured by third-party vendors to provide customized solutions. Arguably, what is “new” about O-RAN is that equipment vendors and network operators are explicit about the coalitions they partner with. Such explicit acknowledgement frames O-RAN as a giant, multi-firm R&D exercise where standards and protocols can be shared widely amongst participants. Interestingly, network operators and equipment vendors appear to be “multi-homing” amongst the various coalitions. AT&T, Comcast, and T-Mobile belong to both the O-RAN Alliance and its “rival” Telecom Infra Project (TIP). (Verizon appears to link only with the O-RAN Alliance. Likewise, traditional vendors do not perceive a rivalry with O-RAN.) Nokia belongs to both alliances, though Ericsson, Samsung, and Siemens belong to the TIP. However, if one alliance achieves technological or usage dominance, the result will likely be a coalescing towards the winning set of standards and protocols.
It is also unclear whether the claimed cost advantages, even if real, are material. While Analysys Mason’s $91 billion annually after 2030 looks like a big number, it reflects global savings potential. Already, $65 billion has been committed in one year alone to boost US broadband infrastructure. The reality is that RAN infrastructure comprises only a very modest component of network operation costs — around 3–4 percent of total costs. By comparison, electricity accounts for 5–7 percent. Even a 30 percent reduction in RAN costs would only reduce the average monthly telecommunications bill by a few cents.
Third, the question of reduced time to market of new innovations is debatable given the extent of testing required to ensure compatibility across an exponentially increasing number of providers — let alone the security testing whenever new componentry is added. The degree of testing necessary for these assurances begs the question of whether the real benefit of traditional procurement arrangements was the single source of responsibility when something went wrong. Who will be responsible in the O-RAN context?
Given that the alliances are not suppliers themselves, risks could be shifted onto operators unless the procurement model does not change much at all, and a small number of firms (albeit maybe including some challengers to those currently prevailing) act as aggregators of supply solutions, as in the past.