“The age of fragmentation of the Internet and of consolidation of its actors”. Interview to Nicola Blefari Melazzi

The situation of the Tlc market and everything that revolves around and derives from it. The future of the Internet, Artificial Intelligence and cellular networks. The growth prospects of 5G and 6G verticals and the consolidation of the European market on the eve of the upcoming elections for the renewal of the European Commission and Parliament. These are the hottest topics in the digital market, at the centre of 5G & Co., the sixth edition of 5G Italy, the leading international conference in the world of telecommunications, organised by the CNIT (National Inter-University Consortium for Telecommunications). We had a chat about it with Nicola Blefari Melazzi, President of CNIT and the RESTART Foundation.

We discussed this with the organiser of the event and CNIT President, Professor Nicola Blefari Melazzi.

Nicola Blefari Melazzi
Nicola Blefari Melazzi

5G & Co. Professor, what is the situation of the Telco sector today?

Nicola Blefari Melazzi. The current situation in the telecommunications market is not rosy, but services and applications based on the 5G network will develop greatly. It remains to be seen by whom they will be implemented and in what manner.

5G & Co. The prospects look good, don’t they?

Nicola Blefari Melazzi. Yes. Of the whole sector, yes. The internet value chain continues to grow at a steady 15% per year and shows no sign of slowing down, while the number of activities that used to be offline and are converted into an online format keeps on growing. However, the market share and revenues of connectivity providers and technology enablers decrease in favour of non-European OTT, hyperscalers and online service providers. The share of revenues generated from advertising is up to 40% of the total earnings of the Internet market, excluding e-retail and e-travel, with very large effects on the economics of the sector (and on privacy)

5G & Co. Are operators losing their edge?

Nicola Blefari Melazzi. Telco companies are not only losing market shares, but they also have decreasing roles in the operation of the network, in the first mile (international links), in the middle mile (distribution), and in the last mile (access).

5G & Co. And how about the OTTs?

Nicola Blefari Melazzi. OTTs are not only “Over The Top” anymore: they are operating also under the top, and that is why it is better to distinguish between real OTTs, which are only content and application providers, and “hyperscalers” that instead also own and/or operate network infrastructures.

5G & Co. What do hyperscalers do specifically?

Nicola Blefari Melazzi. These giants own or operate a large share of intercontinental links and vast areas of the distribution sections, which allows them to reach over 76% of the Internet without traversing Tier-1 and Tier-2 ISPs anymore, more than virtually every other network. Their off-net footprint (servers and functionality installed in the access network of operators) has tripled from 2013 to 2021, reaching 4.5k networks, which provide access to a significant fraction of the end-user population (potentially 50%). In 2023, for the first time, telecoms operators spent more on external cloud and IT providers than on their own analogous in-house services.

5G & Co. What are the effects of this enormous power of hyperscalers on the European market?

Nicola Blefari Melazzi. These trends are global but affect more the EU market, where the number of operators is much greater than in areas with corresponding user populations: Europe counts 45 large mobile operating groups, compared with 8 in the USA, 4 in both China and Japan, and 3 in South Korea4. The European telco sector continues to underperform global peers, both in terms of revenues (decreased in real terms, inflation has been absorbed by operators at the advantage of customers) and in terms of investments. Europe lags all global peers (South Korea, the US, Japan and China) in 5G coverage, usage and performance and in gigabit-capable coverage4, falling short of the ‘full gigabit connectivity’ target. While revenues decreased in real terms, the cost of labour, equipment, materials and energy kept on rising for telcos.

5G & Co. What are the main issues in the European Telco industry?

Nicola Blefari Melazzi. The sector suffers from ancient evils, such as excessive cost competition; “unfair” competition with hyperscalers in terms of regulation; spectrum cost ; growing energy costs; lack of readiness to offer evolved services.

5G & Co. Do you see any consolidation of the EU telcos in the short term?

Nicola Blefari Melazzi. Consolidation of telco operators, which can be a solution to excessive competition, is difficult and painful to achieve and it implies anyhow a transition period, and it is not certain that it will solve their issues, as similar trends are observable in the US and China, where the number of operators is much smaller. And in fact, difficulties in investments are observable also there, e.g., the ranking of deployments of 5G SA, Open RAN and Edge Clouds in Asia, Europe and US are intertwined, depending on which one considers. Now, the issue is not the transformation of an industrial sector per se, which happened several times in the past, with some economists even invoking the need for Schumpeter’s destructive creations. The issue for us is that this specific transformation favours non-EU players and put at risk much needed investments in the European network infrastructure.

5G & Co. Anything else about hyperscalers? Other negative implications?

Nicola Blefari Melazzi. There is also a second concern deriving from the increasing role and size of hyperscalers, and that is that services and applications tend to cluster in parallel, separate ecosystems: we are moving toward a fragmented Internet, with many implications, several of them negative:

  • risks of accumulating power and limiting innovation: we almost made a full circle from old operator monopolies to a distributed environment and now back to concentrated media and platform giants;
  • the intermingling of roles among classic operators, hyperscalers, and other players makes it difficult to coordinate the components of a service and identify responsibilities in case of problems;
  • the same is true for security, privacy, and intellectual property;
  • network measurements and analysis become more and more complex and insufficient;
  • functions such as networked processing, provision of differentiated levels of quality of service, and even basic network services are provided in several cases by means of proprietary solutions whose actual operation is not publicly known;
  • the old digital divide becomes a layered digital service divide: users can find themselves discriminated not only in terms of access and connectivity performance, but also in terms of additional features and services and related performance;
  • spaces for open/academic research outside of collaborations with private companies governed by restrictive non-disclosure agreements are reduced.

5G & Co. Is Internet fragmentation a temporary phenomenon?

Nicola Blefari Melazzi. No, it isn’t. The fragmentation of the Internet and the consolidation of its players will increase with the further expected increase of video traffic and Virtual Reality/Augmented Reality applications and especially of Artificial Intelligence deployments, which will result in, among other things, a limitation of information about how the network works, with increased explainability and accountability problems.

Finally, the trend of network fragmentation and consolidation also manifests itself at the application level (not to mention geopolitical fragmentation): the Internet tends to “channel itself” becoming like the generalist television of the past; less peer-to-peer exchanges between users and search for different content, more concentration, with few platforms that generate content and entertainment and on which the majority of users are focused, and with business models based on advertising.

5G & Co. What can be done to keep an open Internet, not too much TV-like?

Nicola Blefari Melazzi. In this context, it is important to safeguard a plural and open model as much as possible. We must strive to keep the Internet open, standard, globally connected and unfragmented, with access, research and innovation as free as possible. To this end, regulation and provision of specialized and advanced services play a key role. We must avoid getting to the point where the only solution would remain the intervention of antitrust authorities, not least to avoid the negative consequences of such actions, provided we get there.

5G & Co. Let’s talk about smartphones and 5G now. How far have we come?

Nicola Blefari Melazzi. Let us now move the focus from the whole Internet to its cellular section, starting with an empirical consideration: cellular generations unfold their full potential every two generations, meaning that the full adoption of one generation is realized in the next generation. Cellular telephony introduced with 1G has only really spread with 2G. Internet and mobile data use introduced with 3G has only really spread with 4G. Applying the same line of thought to the present case, cloud services, edge-cloud, network slicing, network softwarization and AI, introduced with 5G will fully spread with 6G. Now, in our case, the prediction is perhaps overstated, but it is worth reflecting on the reasons for the apparent pairwise trend of cellular generations.

5G & Co. Tell us more about this pairwise trend of different cellular generations.

Nicola Blefari Melazzi. Radio is a medium and “connects” user devices with other user devices or with servers / clouds and then enables services and applications. The radio is “in the middle of” two other components of the system: “terminals” and “servers”; if adequate user devices or adequate servers / clouds / services / applications are lacking, the radio part has little reason to exist, and, therefore, does not fulfil its full potential. For example, 1G did not have adequate phones that came only with 2G and 3G did not have the evolved smartphones and sufficient mobile applications that came with 4G.

5G & Co. And about 5G? What is still missing?

Nicola Blefari Melazzi. We still lack real innovative devices and significantly new services and applications. The real innovation of 5G lies not so much in the performance improvement of the radio performance, which is certainly important, but which “merely” continues a path of improvements. The real innovation of 5G lies in its softwarized core network, cloud and AI functionality, which only recently is becoming a reality (see the recent establishment of the AI-RAN Alliance.  ).When these become widespread, we will witness an explosion in the use of 5G, which is already starting to happen.

5G & Co. What about radio capabilities, instead? Is 5G keeping its promises?

Nicola Blefari Melazzi. From the point of view of radio capacity, 5G did start to deliver on its promises, by supporting greater capacity and user density: just look at the impressive growth in traffic, both per device and overall, documented in recent technical reports  . As traffic increases, costs to users decrease in relative but also absolute terms. This is both good and bad. Good for consumers and for making sustainable new applications with higher definition video, holograms, virtual reality, AI applications etc.; all these require high transmission rates and the exchange of large amounts of data. These new applications would be unsustainable without a decrease in the cost per transmitted bit. Instead, it is bad for those who must make investments and build infrastructure. Declining market share and revenues of telco operators puts investments at risk and, in the absence of systemic and regulation interventions and new architectures for service offerings, will further shift the balance from telecom operators toward hyperscalers and other players who can invest in infrastructure.

5G & Co. What should be done to cope with this situation?

Nicola Blefari Melazzi.  Four lines of action seem necessary:

  1. Consolidation of operators, with several caveats, though.
  2. Improved regulation to make the sector sustainable, which today it is not.
  3. Evolved value chains. Operators have powerful infrastructures and a wealth of data. They should exploit, beyond connectivity, the ability of offering computing resources, infrastructure elements, duly processed data and insights for the benefit of application providers and of their revenue streams. These aspects could re-establish a prominent role of operators in an evolving environment that will encompass diverse players, e.g., tower companies, system integrators, utilities, municipalities, industries, as well as neutral hosts, private networks and big systemic vertical companies, which are deploying their own communication networks; otherwise, all these actors will try to fill the void left by operators, as they already started to do.
  4. Evolved service offer. Dynamic delivery of tailored offerings to the customers, and maximization of the usability, programmability and exploitation of the infrastructure, through the establishment and wide-spread use of intent-oriented mechanisms and of exposure of network and service APIs. This is key to facilitate the convergence of European electronic communications networks and cloud services, as recommended by many.

5G & Co. What can be done to ensure bespoke offerings for customers?

Nicola Blefari Melazzi. Points 3) and 4) have a basic pre-requirement: the deployment of stand-alone 5G networks. Today, in Europe, the 5G core network and its cloud (and AI) functionality is not significantly implemented; we have (not everywhere!) the radio part of 5G with better performance than 4G, but still attached to the fixed, core network part of 4G, and, therefore, it cannot unfold its full potential. It is like having a restaurant under new management, with new and experienced service staff, but with the same old kitchen. Furthermore, unfortunately, even stand-alone 5G will not be enough to guarantee success; networks have to offer more (i.e., computing resources, data/insight), in a simpler way (intent, APIs, etc.), and both services and applications must be developed accordingly to get the most out of it, with appropriate related business models.

5G & Co. How to get out of this deadlock?

Nicola Blefari Melazzi. Here comes the role of research and development in resolving the situation just outlined; it is the most urgent R&D topic to be addressed in Europe, because, without an economically sustainable sector, which is capable of making investments, all other technical improvements are certainly of interest, but less useful for our countries.

In fact, we already said that 5G provides a significant level of innovation compared to 4G, in terms of softwarised/cloud/intelligent core networks. However, further steps are needed to improve it, also in terms of service offerings and to move towards 6G.

5G & Co. There is no stand-alone 5G yet and we are already thinking about 6G?

Nicola Blefari Melazzi. There is no need to be afraid: we are not talking about new, heavy infrastructure to implement the future 6G. We do not necessarily need new radios and new equipment, only a more advanced software. It is necessary to use an architecture that we can call 6G but which is nothing more than a software evolution of 5G and is not about radio performance. It is an architecture for building and selling customised ICT services for both end users and verticals. Services must leverage on connectivity, context-awareness and cloud and AI capabilities, services which range from basic communication and radio sensing capabilities to high-level processing tasks and network data-driven insight. They must evolve from the current predefined and standardised solutions offered to a broad customer base, to dynamically generated ad-hoc solutions customised to meet specific individual needs. The creation flow should be either bottom-up, designed by the vendor, or top-down, with customers making specific requests, assisted by Artificial Intelligence, which are fulfilled by exploiting available capabilities.

5G & Co. What is the goal of this architecture?

Nicola Blefari Melazzi. The reason behind this architecture is twofold: on the one hand, the aim is to provide customisable, AI-driven, efficient and sustainable services that are directly accessible by end-users and verticals. On the other hand, the goal is to seek to strengthen the position of EU industrial and commercial ICT players.

5G & Co. Can you provide more details? Is it about the development of verticals?

Nicola Blefari Melazzi. Not really. Regarding the first point, this concept evolves the 5G system along four directions:

  • Dynamic, tailored services. From generic, static, one-size-fits-all offerings to unique, tailored and dynamically customizable services. An important example of a scenario where specialized services could be deployed is that of private networks and indoor networks .
  • New high-layer interfaces. From network-level interfaces that require specialized telco knowledge to be setup and used to higher-layer, intent-based interfaces empowered by AI, till natural language interfaces comprehensible to verticals without requiring them to oversee or manage network infrastructures.
  • Resources beyond communications. From predominantly basic connectivity services to elaborate, comprehensive solutions integrating cloud and AI functionality to their fullest extent.
  • Better network interfaces. From network-level interfaces that offer restricted interaction solely with the network control plane to interfaces capable of delivering insights (useful to application developers) that leverage on network-collected data, duly processed in accordance with legal and ethical frameworks . Our system envisages, among others, the exploitation of localization systems and data, of sensing information, of traffic and mobility conditions, of user data across different abstraction levels, managing and transferring content across computational tiers of the cloud-edge continuum, integrating security and risk management measures, and enabling programmable sensing and service coverage. Such capabilities result in a significant expansion of network programmability and service customization.

5G & Co. How could this architecture benefit EU operators?

Nicola Blefari Melazzi. Benefits would include substantial cost reductions, accelerated time-to-market for new services and products, enhanced scalability, adherence to regulatory standards, and fortified security measures. Telcos would be poised to broaden their functions, transcending their traditional role as mere bandwidth conduits. They would have the opportunity to cultivate new revenue streams by providing API access and additional value-added services. Furthermore, by embracing evolved interfaces, telcos would have the opportunity to strengthen their partnerships with cloud service providers and content distribution networks, fostering a more collaborative and efficient ecosystem for joint service innovation and offerings.

5G & Co. What are the advantages for non-telcos and private networks?

Nicola Blefari Melazzi. There would also be benefits for vertical industries, which could more easily deploy sophisticated, multi-regional applications distributed horizontally and vertically, thus removing the need to integrate resources across multiple administrative domains, a challenge so far handled by hyperscalers only.

Given the lack of local cloud and OTT giants within the European Union, as well as the need for telecommunications operators to expand beyond the mere transmission of data, the presence of a beyond 5G/6G network capable of offering specialised, programmable and bespoke services would significantly strengthen the EU’s information and communication technology (ICT) economy.

Finally, an important remark is that more attention must be paid to private network scenarios, whether based on operator-provided or non-operator-provided spectrum, and to indoor environments and gathering places, such as hospitals and transport systems. Business opportunities in these scenarios are significant.