Bank of America
Bank of America: OpenRAN primer with global 5G implications
Written by Multiple Bank of America Research Analysts
OpenRAN turns base stations using proprietary hardware into software running on common off-the-shelf hardware. Tal Liani views this trend as a continuation of ongoing forces in the broader IT and hardware markets – a shift to virtualized hardware, merchant silicon and general software disruption of proprietary hardware markets.
With 78% of the cellular base station/ RAN market controlled by Huawei, Ericsson and Nokia, OpenRAN represents an opportunity for software-only vendors like Altiostar, Parallel Wireless and Mavenir to offer traditional as well as new carriers such as Rakuten and DISH the technology to build cost effective networks with no legacy equipment consideration. OpenRAN is ideally deployed in a virtualized or cloud-based architecture, offering a high degree of automation as well as enabling vendor mixing/matching to reduce costs of innovation and increase efficiencies.
$35bn RAN market set for new competition, disruption:
Radio Access Networks (RANs) represent the largest category of hardware technology, estimated at just under $35bn in 2020. The market is highly concentrated, with 78% controlled by Huawei, Ericsson, and Nokia. OpenRAN is disrupting the market, enabling vendor mixing and matching, greater competition, and introducing new entrants via standardization, separation of software and hardware, and by turning certain elements into independent applications.
Early days for OpenRAN, likely limited impact in near-term:
OpenRAN is still in its early days, representing ~1% of the total RAN market in 2020 and ~6% projected by 2024. These estimates are preliminary though, and like any other disruptive technology, deployments could gain momentum as they are adopted by key carriers. To date, the biggest OpenRAN advocates are new carriers, like Rakuten and DISH, which are utilizing the technology to build cost effective networks having no legacy equipment considerations. Tier-1 carriers are taking a more measured approach with legacy architectures/vendors still offering better economics. Yet, at the same time, most leading carriers are testing OpenRAN and some already deploy it on a small scale. For example, Verizon‘s 5G upgrade is partially done with Ericsson‘s proprietary approach and partially with Samsung‘s OpenRAN solution. We also see interest from European carriers, with OpenRAN providing an efficient way to replace Huawei equipment.
Importance of semis, commodity hardware on the rise:
On the vendor side, OpenRAN represents new opportunities for software-only vendors like Altiostar, Parallel Wireless, and Mavenir, hardware providers like Fujitsu and leading semi vendors like Intel and Qualcomm. Nokia and Ericsson partially support OpenRAN, given its disruptive nature, focusing instead on proprietary software-based solutions, like virtualized (vRAN) or Cloud RAN. However, as momentum grows, we expect all leading vendors to support OpenRAN, similar to trends seen in Switching and Routing. We expect the legacy Radio vendors to offset the negative implications of OpenRAN via a growing focus on software, applications and expansion into adjacent markets.
Open Radio Access Networks, or OpenRAN, is an emerging trend that is set to shake up the roughly $40bn 4G/5G infrastructure market. We view this trend as a continuation of ongoing forces in the broader IT and Telecom Hardware markets, such as the shift to virtualized software, whitebox hardware, merchant silicon, and general software disruption of the proprietary hardware markets. We have witnessed similar efforts to open and standardize other networking markets such as Ethernet Switching and IP Routing, however, the complexity, performance demands, and tight vendor controls in the Mobile Infrastructure market have left the Radio Access Networks proprietary thus far. 5G deployments represent an entry point and a catalyst for OpenRAN, and our deep dive aims to explore the potential opportunities and disruption across vendors and sectors that service the RAN market.
Our note is organized into 5 main sections: 1) drivers of OpenRAN deployments vs the challenges, 2) introduction to the Radio Access Market, which describes the technical components, leading vendors, and market dynamics of the traditional RAN market, 3) an OpenRAN 101 section that outlines the architectural changes, new vendors, and growth forecasts, 4) the impact of OpenRAN, and opportunities related to adjacent areas of technology, such as semiconductors, and lastly 5) OpenRAN traction by geographic region.
Ultimately, we see three key takeaways for investors: 1) we flag that OpenRAN and Virtualized/Cloud RAN are separate trends that are coming together to form the attractiveness of Open and Virtualized RAN (vRAN), 2) it remains early for OpenRAN, which is expected to represent less than 1% of RAN spending in 2020 and grow to only 6% of the total market by 2024, and 3) OpenRAN creates opportunities for legacy and new vendors, in our view. We quantify the disruptive potential on page 33, and compare the OpenRAN phenomenon to what happened in the virtualized Evolved Pack Core (EPC) market. We also view OpenRAN as an enabling technology for global carriers to replace Huawei in certain areas, either through growing share of Nokia and Ericsson, or via the introduction of new vendors.
Figure 1: OpenRAN vs Virtual RAN and evolution to Open vRAN
Source: BofA Global Research, Omdia
A key benefit of OpenRAN is new innovation in the radio access market, with newer software companies such as Altiostar pushing incumbent vendors to begin disaggregating their software from hardware and support industry groups dedicated to developing OpenRAN technology, such as the O-RAN Alliance. OpenRAN itself also encourages innovation via its open interfaces, and the enablement of third party vendors to add new solutions. OpenRAN is ideally deployed in a virtualized and cloud-based architecture, offering a high degree of automation, increasing efficiencies and reducing costs. Currently, OpenRAN development is supported by a wide range of semiconductor, hardware, testing, systems integration, and software companies, helping foster innovation in each domain and cooperation toward a more ‘hyperscale-like‘ network. As we have seen in other areas of cloud networking and technology, open ecosystems often foster greater innovation.
Figure 2: Comparison of Traditional RAN versus OpenRAN
|Source: World Wide Technology|
Less vendor lock-in to create more competition:
Another key driver for OpenRAN interest is the ability to avoid vendor lock-in. Following years of vendor consolidation in the Mobile Infrastructure market, there are only four leading equipment provider choices: Huawei, Nokia, Ericsson, and ZTE to a lesser extent. On top of limited choice, it is notoriously difficult to switch vendors, requiring expensive and labor-intensive equipment swaps from the radio head to the baseband data center infrastructure. In some cases, the equipment swap cost burden falls on the carrier and in some cases vendors provide such services as part of the sales/services strategies. Ultimately, the lack of choice and difficulty in switching vendors create a market rife with equipment vendor lock-in.
Swapping out Huawei represents a major catalyst:
As the market has consolidated, political pressures versus Chinese vendors‘ role in 5G (see note on Huawei pressure and risk) further limit vendor choice to only 2-3 firms in some regions. Therefore, global pushback against Huawei/ZTE may be one of the largest drivers of OpenRAN adoption, pulling forward the timing of operator decisions on RAN architectures. Huawei has gained significant share in the $38bn market over the last seven years, now representing 34% of the total market, and government support for removing the vendor from networks has grown in recent months. The UK government recently instituted a policy banning UK carriers from buying new Huawei equipment beginning in 2021, and all Huawei equipment must be removed from UK networks by 2027. Other regions of Europe such as Belgium, Poland, and Sweden have also recently shied away from Huawei.
Importantly, replacing Huawei brings large costs, both from losing Huawei as a competitor (Huawei known for its attractive price/performance) and equipment swaps. As a result, the US government is beginning to take steps to help developing countries within Africa and the Middle East fund the costly replacement of Huawei/traditional equipment. Specifically, the US Agency for International Development is spearheading the effort, while the US State Department continues to pressure US allies to displace Huawei and ZTE equipment from their networks. In our view, the replacement of Chinese RAN technology could open up a $35bn market to both incumbent and new vendors, and the replacement of network vendors‘ architectures offers an attractive opportunity for carries to re-architect the Access network utilizing modernization and virtualization, which are both drivers for OpenRAN. The US government has also explored investing in OpenRAN technologies to help US software/hardware/semi vendors play more of a role in cellular networks.
Open RAN Cost Savings:
Opening the interfaces between the baseband unit (BBU) and remote radio unit (RRU) helps increase competition, lowers the switching costs, and likely saves carrier capex to some degree. However, we believe the real benefits related to the OpenRAN vision come to fruition when the architecture becomes virtualized or cloud OpenRAN (often referred to as Open vRAN). In Open vRAN, carriers first save on equipment capex as the baseband unit software runs on commodity off the shelf (COTS) hardware (i.e., x86 servers) rather than proprietary integrated hardware. Software can be purchased from new vendors and the equipment can be provided by vendors such as Quanta Computer. High degree of competition for the RRU component and the hardware commoditization for the BBU component could result in potential capex savings of 40-50%. Installation and integration services can also potentially be brought in house or outsourced to a longer list of competitors, adding RAN installation savings that are typically part of capex (see Exhibit 3).
The second area of carrier total cost of ownership (TCO) savings is related to the maintenance and operating expense. By copying the efficient cloud models of hyperscalers and centralizing/standardizing the foundation of the RAN, carriers stand to run more efficient data center operations. The software-defined approach also adds to network agility and automation. Through better agility and automation, carriers save on the management, maintenance, and upgrades for the network. Early reports suggest potential 31% operating expense savings as a result (see Exhibit 4).
OpenRAN Industry groups:
1. Telecom Infra Project (TIP):
TIP was formed in February 2016, with Facebook playing a central role. Major vendor Nokia is also part of the group. Japanese members include NTT, KDDI, SoftBank, Rakuten Mobile, NEC and Fujitsu. TIP‘s goal is to create mobile networks using open and disaggregated solutions. The scope of the group‘s work extends from OpenRAN to include the backhaul portion of the network, core network architecture and other areas.
TIP and O-RAN Alliance announced a liaison agreement in February 2020. Under the agreement, the groups will share information, reference respective specifications and collaborate on testing.
2. O-RAN Alliance:
O-RAN Alliance was formed in February 2018 through a merger between x-RAN Forum and C-RAN Alliance. Japanese members include NTT DoCoMo, KDDI, SoftBank, Fujitsu and NEC. In May 2020, O-RAN Alliance and GSMA, an industry body representing MNOs, agreed to collaborate on opening up 5G networks.
Major vendors Ericsson and Nokia are also part of the O-RAN Alliance. However, they appear to be taking a slightly different stance on OpenRAN. In February 2019, some members of O-RAN Alliance announced new open fronthaul specifications and related testing. Nokia is mentioned in the announcement, but Ericsson is absent. Nokia believes open standards are a viable option for RU-DU, but is doubtful about the effectiveness for the CU-DU interface.
The O-RAN Alliance is focused on efforts to standardize technologies. NTT DoCoMo is expected to play a key role in the standardization process. In September 2019, DoCoMo announced it had achieved interoperability with equipment from different vendors in a 4G/5G demo project. Vendors in the demo were Nokia, Fujitsu and NEC. As widely reported, Fujitsu has teamed up with Ericsson and NEC is collaborating with Samsung in OpenRAN technology.