Article  | 

The gRAN is proposed as an alternative to the traditional link network

SHARE

Virtual link networks (vRAN) were presented as an alternative to traditional link networks (RAN) towards the middle of this decade, at least in reduced environments, such as in private corporate networks or on the edge of telecommunications networks. The open and unbundled link networks, known generically as Open RAN, will have to wait a few years for the technology to mature, according to the opinion of some experts and specialized consultants. The cloudification and virtualization of telecommunications networks (increasing use of the cloud and remote computers) is seen as an unstoppable process, but that does not necessarily mean that the networks are open or disaggregated.

The commercial and technological impact that open (Open RAN) and virtualized (vRAN) link networks will have in the coming years has been causing intense debate in the mobile phone industry lately, with the emergence of alternative providers. There are also discussions at the political level, especially in the United States, because they want to find a way to achieve greater technological sovereignty in the development and construction of equipment.

Unlike previous generations of mobile telephony, which were primarily aimed at consumers and companies ‘voice and data communications, 5G networks will form an essential part of companies’ technology, manufacturing and commercial platforms, both in internal communications as well as in relationships with suppliers and customers, especially during the second half of this decade, when 5G networks are at their peak.

Virtual link networks will require the development of highly sophisticated cloud platforms to manage the flow of information and the necessary flexibility, says Analysis Mason.

Although the repercussions on the 5G network industry and its repercussions were well known, the pandemic has clearly revealed its great strategic importance for the coming years and the great dependence of the United States on a handful of manufacturers of proprietary network equipment. . In practice, three companies account for more than 70% of the global market for link networks (Huawei, Ericsson and Nokia) and another three that want to gain ground (Samsung, ZTE and NTT DoComo, the latter being supported by the also Japanese Fujitsu and NEC ). None of them are American; Cisco and Ciena have a small share of the market for networking equipment, but not for networking.

The shortage of semiconductors and the breakdown of the global logistics chain has revealed the magnitude of the dependence on the United States, which continues to be the undisputed world power in the field of software, including chip design, but does not have a large manufacturing industry chip is not in the lead. The most innovative American companies must send their chip designs to a small island next to China for TSMC to manufacture them.

Ironically, the semiconductor shortage has been exacerbated by the embargo imposed by Donald Trump on the supply of chips for smartphones (the Kirin) that TSMC made to Huawei. Huawei’s smartphone business has sunk as a result of the embargo and US pressure not to buy Huawei network equipment from European operators also puts the Chinese company in serious trouble but it has not been resolved, it has even worsened. what the United States wanted to solve: regain sovereignty in the manufacture of chips.

Now the United States wants to recover lost ground in a hurry with an aid of 50,000 million dollars to boost the manufacture of chips by US companies, an initiative that in any case will take several years to bear fruit. For now, what it has done is to further stimulate China’s desire for technological sovereignty, especially in the manufacture of chips, as was already revealed in 2015 in its Made in China 2025 program, which set off all the alarms in United States but that, apart from turning the entire global chip supply chain upside down, has had few practical results.

The Biden Administration has maintained the diplomatic and commercial pressure against China that its predecessor initiated, thus for the moment it has only increased the tension between the United States and China. Now, on the subject of link networks, what the United States intends is to promote the development of fully open and disaggregated link networks, which is known as the Open RAN initiative, and designed mainly by small American companies, such as Mavenir, Altiostar or Parallel Wireless, among others, against the European and Chinese colossi.

The main objective of the United States is not that its companies have a large participation in the manufacture of 5G networks, which it is aware that is a chimera, but that it is preparing to be someone in the development and subsequent manufacture of the networks of the next decade , 6G, an objective that is also very complicated because, apart from the fact that European and Chinese companies want to maintain their leadership position, mainly by asserting their immense portfolio of patents in mobile phone networks, Japan and South Korea pursue the same objectives . The struggle to set specific standards in 6G networks, which has already started, is expected to be colossal.

Building vRAN, a first step

The United States is not the only one interested, for reasons of technological sovereignty, in having more diversified and disaggregated mobile access networks, so that there are multiple suppliers of standardized and interchangeable components. “Mobile phone operators, motivated by the convergence of technical, economic and political factors, are promoting the design of new network architectures based on the principles of virtualization, cloud computing, open interfaces and network automation”, as Mani Manimohan, responsible for policy and regulation of the digital infrastructure of the GSMA, the association that brings together the world’s major telecommunications operators, says.

But, as Manimohan points out, “open interfaces and virtualization are not technological principles by themselves; it is the application of these principles, with the disaggregation of the infrastructure of the link networks into interoperable elements of hardware and software, which is creating alternative paths for the construction of mobile networks ”. Precisely to favor and promote the development of supply alternatives, the large operators recently signed a memorandum and also call for more flexible regulation to face the challenge of 5G.

“Open and virtualized link networks (Open RAN and vRAN) are expected to deliver technical and economic benefits by increasing the flexibility and diversity of network equipment and the software ecosystem, as well as fostering innovation,” says Manimohan. “By reducing cost, Open RAN and vRAN will also help optimize the nearly $ 1 trillion investment that operators around the world will need to make to achieve their proposed connectivity goals.”

Enterprise private networks can be a conducive use case for vRAN development, especially for single vendor trunks, giving way to open vRANs

However, the GSMA executive warns with finesse, “it will be necessary to solve multiple challenges to achieve the massive deployment of these technological principles” on open and virtualized link networks. The GSMA published last February a guide for politicians on Open RAN and vRAN, in which it explains their advantages, disadvantages and challenges to overcome

Diversifying the supply chain of the link networks and ensuring that the networks are reliable and secure is causing many governments to take sides and have decided that some suppliers considered high risk cannot have their equipment installed in the networks of some countries, such as This is the case of the United Kingdom, which has forced its operators to remove Huawei equipment from their networks before 2027, especially in their backbone networks.

These decisions are creating strong controversy within some countries, fearing reprisals from China. Germany, and especially Angela Merkel, Chancellor until September, is the one that is the most reticent, because her country depends a lot on exports of machinery and cars to China. Sweden decided to prohibit Huawei from participating in a recent competition for the supply of 5G network equipment, despite the fact that Börje Ekholm, CEO of Ericsson, a Swedish company, warned of the dangers posed by this measure and feared that it would be she who would end up paying the costs. broken plates.

This same week, the Global Times newspaper, owned by the Chinese Communist Party, published a series of articles on the advisability of Ericsson continuing to supply 5G network equipment in China if Huawei is denied access to the Swedish market. The newspaper, which is taken for granted that he has an official opinion and with the approval of the Chinese Government, invites Ericsson to go to China to “interview” about the opportunity for the Swedish company to access the Chinese telecommunications market .

Need for a robust cloud for vRAN

The experts, despite Mani Manimohan’s statements, are not clear about the commercial path that virtualized access networks can take, beyond their use in specific and reduced environments, much less truly open access networks. Its development and manufacture on an industrial scale would mean overcoming numerous technical obstacles and a considerable investment, difficult for companies with a size and resources much smaller than those they want to face, apart from the fact that the latter could also develop their own open networks and sweep away potential competition.

In an article from the end of January, the consulting firm Analysis Mason exposed the difficulties for the development of vRAN that operators will have to face. The consultancy’s central argument to justify its skepticism is that the virtualization of link networks requires the development of a new cloud platform.

The operators’ wishes are clear: they want software that runs the link networks on a platform located in the cloud to have the agility and the reduction of costs derived from the automated use of a native cloud, a standard hardware and the flexibility of resource scalability. With this strategy of an “open” and horizontal RAN that is managed through a native cloud, the operators could select the functions of the RAN that best suit their interests and not depend as now on proprietary teams on a monolithic network , as Analysis Mason points out.

The main problem, in the opinion of the consultancy, is that the functions of a 5G link network are highly complex, and much more when massive MIMO antennas are generalized to understand the coverage and capacity to capture signals from a station. mobile link.

These complex RAN functions, together with the need to provide excellent performance and with very low latency, will require a highly sophisticated cloud platform capable of handling these demands, and probably build a network infrastructure that links the virtualized RAN with a new generation of cloud platforms. In addition, this cloud should be able to manage multiple vRANs, in a remote centralized environment.

Aside from the sophisticated cloud, Analysis Mason ensures that a RAN cannot be fully virtualized. In the opinion of his analysts, the link network is the last bastion of a telecommunications network that can be virtualized. A RAN has three basic layers and six essential functions, some of which must necessarily be next to the antenna, because the latency must be minimal, practically instantaneous.

The information process in a 5G RAN is much higher than in a 4G, due to the higher bandwidth with 100 MHz channels instead of the 20 MHz of a 4G RAN, since massive MIMO antennas require 20 times more capacity of baseband processing and transmission times of 0.2 milliseconds versus three milliseconds for 4G.

vRAN on private networks

These constraints mean that operators must build cloud-based platforms capable of supporting highly distributed link networks, with very low latency and high performance, which will necessarily be very expensive. This will in turn require new use cases that bring additional revenue to operators and justify and monetize the highly sophisticated cloud.

Utilities Spending at the Enterprise Edge vs. Total

Source: Analysis Mason, 2021

Throughout this decade, enterprise spending on edge cloud versus total edge public spending is projected to increase, from just over 5% this year to 15% in 2023 and close to 35% in 2030, as seen in the graph below. This increased investment by companies in edge clouds will increase the interest of operators to provide virtual link networks because they will have more revenue and will allow these more sophisticated clouds to be deployed.

Growth forecasts for virtual link networks, including open ones, are relatively moderate for the next few years. According to the consulting firm Omdia, in 2024 10% of the link networks will be open and virtualized, as indicated in the graph below, although it is certain that as the decade progresses this percentage will be much higher.

In a later article from the end of March, Analysis Mason continues to be cautious about the development of virtualized networking, especially from centralized cloud environments, in the next two to three years, due to the low technological maturity in this area. However, in corporate private networks, where the needs are more specific and concrete, the construction of vRAN is much more feasible, because the complexity of the cloud that supports them will also necessarily be a lot. Therefore, there is a good future for corporate vRANs in the medium term.

Virtual link networks are therefore a significant segment of future investment for operators and, as Analysis Mason adds, will pave the way for open link networks. The consultancy’s forecast is for operators to start building virtual link networks for companies with very specific requirements and for a single RAN manufacturer. Over time, vRANs will support more complex environments and will subsequently be open, unbundled, and multi-vendor, already in the second half of this decade.