Home Blog Page 2

Vodafone deploys conical 5G antennas to hit dead spots in tunnels


Road and rail tunnels can be difficult areas to cover with 5G so the operator has tried some new multi-band antennas to help out 

Vodafone Germany has commissioned a new type of conical multi-band mobile antenna for more stable 5G coverage in tunnels. The operator has deployed five of these antennas in the 1,400-meter-long Arlinger Tunnel near Pforzheim, improving mobile reception. Ericsson’s mobile antenna supports multiple frequency bands and is specifically designed for use in tunnels.

Interestingly, the antennas are cone-shaped and the form factor makes it particularly wind-resistant and therefore robust, according to the operator. The antenna covers all low band and midband spectrum for all mobile generations and is quite unique with respect to the 2×2 MIMO capability (XPol) which is required for 5G services. They also transmit multiple 5G and LTE frequencies simultaneously, reducing the need for multiple antenna types and improving efficiency.

“Closing dead spots in tunnels is particularly challenging for structural reasons alone. Passing cars and trains set large air masses in motion in tunnels, which can cause vibrations in the antenna technology and thus impair the transmission and reception performance of mobile phone antennas,” said Vodafone head of network development Marc Hoelzer. “This places special demands on the wind resistance of the antennas used and their installation.”

He added: “For drivers and passengers, the commissioning of the new antenna generation will result in fewer dropped calls and more stable data rates.”

In Germany, there are over 270 road tunnels with a total length of 270 km on federal highways. There are 420 tunnels on rural, district, and urban roads with a total length of over 350 km. The Deutsche Bahn rail network comprises 761 tunnels with a total length of more than 600 km. Vodafone plans to use the wind-resistant conical antenna in construction projects in an initial 20 additional tunnels.

In smaller tunnels, mobile phone antennas at the entrances and exits are often sufficient for mobile phone reception along the tunnel route. In deeper and longer tunnels, optical amplifiers (repeaters) are used, which transmit the mobile phone signal to numerous tunnel antennas installed in the tunnel and then return the signal to the mobile phone cell at the end. 

Implementation also presents many challenges. Cables must be laid and technology installed. This usually requires the tunnels to be closed by Deutsche Bahn or the Autobahn GmbH. Therefore, network expansion in mobile phone tunnels usually takes place as part of renovation or maintenance work. 

To avoid every mobile network operator having to install its own cables and technology, there is usually a responsible network operator for each tunnel and the other providers connect their technology to the infrastructure.

Boldyn’s hold-my-beer

Overcoming the unique radio propagation characteristics in tunnels has been a conundrum for operators and vendors alike. In places like the Nordics and Australia, it has resulted in some interesting developments in underground mines. 

While Vodafone’s deployment of this antennas is helping over what you could argue is shorter distances, spare a thought for Boldyn Networks which signed a 20-year partnership with Transport for London (TfL) to provide high-speed mobile coverage across the London Underground network. Once done, customers of all the UK’s mobile network operators (MNOs) Three UK, EE, Vodafone and Virgin Media O2, will be able to access high-speed mobile connectivity while travelling on the Tube.

To overcome the lengths and twists of the Tube, specially designed radio units will be installed throughout the stations and concourses to beam the cellular signal to phones. Meanwhile, a 175mm-diameter radiating cable is installed on the tunnel walls at train window height to provide signal to passing Tube services. 

This is known as a leaky feeder as it has gaps or slots in its outer conductor to allow the radio signal to leak into or out of the cable along its entire length – both are compliant with current regulations. For longer tunnels, a high-power radio unit is installed every 500m to amplify the signals. 

Orange initiates in-person listening to customers in 17 MEA countries

The operator despatched 10,000 employees from all functions to meet 15,000 customers in 120 cities and villages, on the same day, to gain a better understanding of what they want

Orange has launched an initiative called My Customer, My Boss simultaneously in its 17 subsidiaries in Middle East and Africa. The intention is to improve listening to customers at all levels of the organisation, by sending more than 10,000 employees from all functions to meet 15,000 customers in 120 cities and villages, on the same day.

The initiative was piloted in Sierra Leone and, Orange says, “marks a new stage in the management culture of Orange Middle East and Africa, making every employee an actor of customer satisfaction, whatever their role in the company”.

Employees listened to customers in large cities and more remote areas, including individuals and companies, Orange Money users and partners “to better understand their expectations and uses”.

The initiative is set to become an annual event. How successful it proves to be depends on how well the telco acts on the feedback it garnered, which will, of course, be analysed using AI and “transformed into concrete actions during internal hackathons organized in each country”.

There’s an app for that

The Gofiled app was used for data collection, developed by a startup from Orange Digital Centre in Tunisia. According to the press statemet, this “collaboration illustrates Orange’s desire to combine local innovation, social impact and business performance, by mobilizing its entrepreneurial ecosystem to meet the needs of the field”.

Brelotte Ba, Deputy CEO of Orange Middle East and Africa, comments, “Customer experience is everyone’s business, every employee, without exception, is committed to meeting our customers’ expectations with excellence.

“With My Customer, My Boss, we demonstrate that our commitment to serve is collective, concrete and forward-looking by organizing for the first time, on a continent scale, a collective mobilization of this magnitude. I would like to thank all the employees who volunteered to meet our customers and partners with professionalism, listening and enthusiasm.”

Starlink stalls in South Africa despite progress elsewhere on the continent

Regulations in Elon Musk’s native country insist overseas firms must have 30% South African ownership, which the US-based billionaire says are “a barrier”

Starlink, the satellite-based internet service created and run by the world’s richest man, Elon Musk, has so far failed to strike a deal in South Africa where he was born. Starlink’s parent company is SpaceX and South Africa is the continent’s most industrially advanced nation. Its President, Cyril Ramaphosa, also seems keen to strike a deal, meeting Musk in New York last year.

The sticking point is that South African regulation stipulates that an oversea company must grant 30% ownership by groups that have been historically disadvantaged, which overwhelmingly means South Africa’s black population. In March Musk claimed via his social media platform X, “Starlink can’t get a license to operate in South Africa simply because I’m not black”.

This has been rebutted, most notably by Clayson Monyela, a spokesperson for South Africa’s foreign ministry, who replied in a post also on X, “Starlink is welcome to operate in South Africa provided there’s compliance with local laws. This is a global international trade and investment principle. There are over 600 US companies investing and operating in South Africa, all complying and thriving. Microsoft just announced additional investments yesterday.”

Success elsewhere in Africa

Nigeria was the first African country to grant Starlink an operating licence in January 2023 and it now operates in about 20 African countries, having recently added Guinea-Bissau, Somalia and, earlier this month, Lesotho.

Lesotho granted Starlink a 10-year licence to operate a satellite network and provide satellite internet services. This was in the wake of Lesotho being hit by a 50% tariff on imports to the US by President Donald Trump. The President backtracked and said the tariff would be lowered on a temporary basis.

On 5 March 5, the US President suggested in a speech to Congress that Lesotho is a country “no-one has heard of”. SpaceX submitted its application for an operating licence in Lesotho a year ago but it was opposed by a number of civil activist groups. One of their chief objections was relying on a service that was entirely owned by an overseas company.

Aljazeera has drawn a parallel with another larger-than-life South African, Joshua Norton, who was determined to be part of the US’ story. In 1859 Norton declared himself “Emperor of the United States”. He made zero progress in his bid to assume power, but he didn’t provoke a backlash either. Instead “he was way ahead of his time on human rights issues,” according to John Lumea, founder of the Emperor Norton Trust.

Smartphone shipments stutter as tariff impacts loom


The tech sector shock looks set to rumble on as supply chains realign and European companies rethink cloud suppliers

Global smartphone shipments in the first quarter of 2025 grew 3% YoY, according to preliminary data from Counterpoint Research’s Market Monitor service. The global smartphone market, which grew in 2024 after a dip in 2023, started off on a positive note in 2025 as well, driven by growth in markets like China, Latin America and Southeast Asia. 

However, the US Administration’s capricious will-they-won’t-they approach to replacing global trade with protectionist tariffs could already be having an effect on the handset market killing off any big recovery. Already, Nvidia has headed off to Beijing as the unintended consequences grow.

“The Q1 growth fell short of our earlier projection of 6% as uncertainty started building up around tariffs towards the end of the quarter and as a cautious inventory buildup was seen at OEMs,” said Counterpoint senior analyst Yang Wang. “Uncertainty still looms in the market, which will affect the situation going forward. As a result, our previous forecast of 4% growth in 2025 seems hard to achieve. We may even see zero to negative growth this year.”

Taiwanese analysts TrendForce revealed today that the implementation of the US “reciprocal tariffs” on 9 April – followed by a 90-day grace period for most regions – has prompted semiconductor memory chip buyers and suppliers to adjust their strategies in response to policy uncertainty. Senior research vice president Avril Wu noted that with both sides rushing to complete transactions and shipments within the grace period to mitigate future policy risks, memory market activity is expected to pick up notably in 2Q25.

She said the grace period has temporarily eased concerns about demand loss due to new tariff barriers. However, lingering uncertainty over the direction of US trade policy has driven memory buyers to adopt a more defensive stance—actively raising DRAM and NAND Flash inventory levels as a buffer against supply risk. 

Wu concluded that ultimately, the future course of US tariffs will be the key factor shaping memory supply-demand dynamics and pricing trends in the second half of the year. Needless to say, anyone in the tech industry could substitute the word memory with whatever flavour of tech the US policies will impact this year.

Clouds on the horizon

In the latest FAFO news, OVHcloud CEO Benjamin Revcolevschi said, in a quarterly earnings statement: “In the current geopolitical context, we are seeing a shift in the concerns of private companies and public organisations in Europe. Questions of strategic autonomy are now on CEOs’ agendas.”

He added: “The choice of a cloud provider is no longer just a technical matter, but also a strategic issue.” The sentiment will no doubt be noticed by the three US-based hyperscalers: Amazon Web Services (AWS), Microsoft Azure and Google Cloud, which dominate the market. OVHcloud competitor Iliad, via its subsidiary OpCore that operates the group’s 13 data centres, is set to invest €3 billion in AI infrastructure and data centres. With OVHcloud pointing out the obvious, the next phase in Europe’s data centre investments may involve a dose of schadenfreude-as-a-service. 

Back at the global handset market

Outside the whacky world of tariffs, Counterpoint said the global smartphone market’s shipments grew 3% YoY during Q1 2025, driven by growth in markets like China, Latin America and Southeast Asia. Samsung secured first place, beating Apple by a narrow margin. Unsurprisingly, Huawei was the fastest-growing brand due to its strong growth in the Chinese market.

Based on preliminary data, Xiaomi continued its momentum and held on to its third place with 2% YoY growth. This performance largely came from a strong comeback in its home market China, supported by the brand’s retail and portfolio expansion. OPPO and vivo captured fourth and fifth positions respectively, separated by a narrow margin. Both brands gained from their strong performance in the mid-tier price band and from government subsidies in China.

NVIDIA’s CEO accepts invitation to Beijing after billion dollar losses

The fallout from weaponising tariffs and trade restrictions is a series of unexpected consequences, which don’t look likely to benefit the US ultimately

Jensen Huang, CEO of NVIDIA (pictured) which is one of the world’s most valuable companies, has made a well-publicised trip to Beijing after losing billions his company’s lost billions.

Note that Huang’s firm accounts for the lion’s share of AI chips, which ever assessment you believe: reports from Mizuho Securities the investment bank, stated Nvidia’s AI accelerators account for between 70% and 95% of the market for AI chips. Tech Insights reckons it ‘owns’ 65% of the data centre chip market.

Images of Huang meeting tech leaders, including DeepSeek’s founder, and China’s Vice-Premier, are a refreshing contrast with more usual ones of the coterie of US tech billionaires acting like a Trump fan club. Reportedly, Huang discussed new chip designs to meet Chinese customers’ needs during his visit.

Sudden changes damage commerce

NVIDIA’s multi-billion losses arise from turmoil in stock markets, triggered by President Trump’s tariff ‘policy’ and the escalating trade war with China. A new, surprise restriction prevents NVIDIA from exporting its H20 chip to China – the H20 was designed specifically to accommodate previous trade restrictions by the US Administration. Also, it seemed the H20 would remain exempt from restrictions after Huang had a meeting with Trump earlier this month.

On Tuesday, NVIDIA warned that it could lose as much as $5.5 billion in earnings as a result of the restrictions that ban it from selling to major Chinese clients such as Alibaba, Tencent and ByteDance. Last year NVIDIA’s sales in China were $17 billion.

Nigel Green, CEO of deVere Group, an independent financial advisory organisation, stated Trump’s actions are “a masterclass in the law of unintended consequences”. He added, “This is a stark example of how the current US trade stance is pushing countries and companies further toward China, not away from it – financially, economically, politically and diplomatically.

“Trump’s aggressive and often unpredictable trade policies are eroding trust, triggering realignments, and accelerating the shift to a multipolar economic order that is likely to increasingly sidelines the US.”

Dropping a brick

In response to this unpredictability, countries are diversifying reserve holdings, expanding local settlement networks and establishing bilateral trade deals to bypass the dollar.

According to deVere, China’s yuan is being adopted in more cross-border transactions and the BRICS nations are increasingly settling trade outside the US dollar. BRICS is an intergovernmental organisation comprising 10 countries: Brazil, China, Egypt, Ethiopia, India, Indonesia, Iran, Russia, South Africa and the United Arab Emirates.

Collectively BRICS represents around 45% of the world’s population and 35% of global GDP (based on purchasing power parity) and it seems likely more countries will join. deVere found Brazil and Saudi Arabia, which is yet to formally accept its invitation to join BRICS, have expanded their holdings of Chinese currency

Green claims, “By weaponising trade controls and blindsiding key US firms, the administration is encouraging global actors to build parallel systems and deepen ties with Beijing. We’re seeing a steady, serious pivot. Business leaders are making pragmatic decisions to maintain access to the world’s second-largest economy – with or without Washington’s blessing.

“Jensen Huang showing up in Beijing with a smile, just after taking a multibillion-dollar hit from a White House directive, speaks volumes. It’s not defiance – it’s realism. Business is being done where there’s opportunity and continuity. Right now, that’s increasingly outside the US sphere,” Green concluded.

New Asian powerhouse?

Due to the West’s preoccupation with the Trump Administration’s antics, the meeting on 22 March between top diplomats from Japan, China, South Korea attracted less attention that it would otherwise have done.

As Reuters said, they seeking common ground on East Asian security and economic issues amid escalating global uncertainty, although there is also some more localised horse trading going on. While China said the countries together can exert international influence, Japan asked China to lift its ban on Japanese food imports, and Seoul and Tokyo asked for Beijing’s help to ‘denuclearise’ North Korea.

“Given the increasingly severe international situation, I believe we may truly be at a turning point in history,” Japanese Foreign Minister Takeshi Iwaya said at the start of the meeting in Tokyo with Chinese Foreign Minister Wang Yi and South Korean Foreign Minister Cho Tae-yul.

The journey towards fully autonomous networks

Partner content: Celfocus follows TM Forum’s model and Gartner’s guidelines to help CSPs reach Levels 4 and 5, integrating AI and automation for better customer experience and operational efficiency

Communication Service Providers (CSPs) have been undergoing a paradigm shift towards autonomous networks, aiming to increase operational efficiency, service reliability and cost-effectiveness. TM Forum’s Autonomous Networks (AN) framework provides a structured model for this evolution, categorising automation maturity into five levels – from basic manual operations (Level 0) to fully autonomous and self-governing networks (Level 5).

However, the transition to higher levels of autonomy is not only a technological challenge, but it also requires a strategic transformation of Operations Support Systems (OSS). As Gartner highlights in its research on The 4 Phases of OSS Transformation for Autonomous Network Operations, achieving full autonomy demands a multi-phase approach, addressing system modernisation, AI integration, and closed-loop automation.

TM Forum’s Autonomous Networks Levels

TM Forum defines five levels of autonomous networks:

  • Level 0 – Manual Operations: Processes are fully manual, therefore human intervention is required for all network management tasks.
  • Level 1 – Assisted Operations: Basic automation is introduced, but humans are still responsible for all important decisions.
  • Level 2 – Partial Autonomous Operations: AI and automation begin handling repetitive tasks, reducing human workload.
  • Level 3 – Conditional Autonomous Operations: The network predicts and resolves certain issues with minimal human oversight.
  • Level 4 – High Autonomous Operations: Self-optimising capabilities emerge, but human intervention is still available as a fallback.
  • Level 5 – Full Autonomy: The network operates independently, self-configuring, self-optimising, and self-healing in real-time.

For many CSPs, the goal is to achieve Level 4 or 5. Nevertheless, Gartner’s analysis states that most of them are still in Level 1 or 2, struggling with fragmented legacy systems and limited AI readiness. To progress, CSPs must rethink their OSS strategies, following a structured transformation process.

Gartner’s Model for OSS Evolution

Gartner outlines a four-phase transformation model for evolving OSS into Autonomous Operations Systems (AOS), a critical step in achieving network autonomy.

Phase 1: Clean-up and Modernisation

In this phase, before automation can scale, legacy OSS systems must be streamlined.

It involves:

  • Eliminating redundant applications and consolidating OSS functionalities.
  • Upgrading outdated systems to cloud-native architectures.
  • Creating a target OSS blueprint that enables closed-loop automation at multiple levels (network, service, and order).

For TM Forum, this phase is essential for CSPs moving from Level 1 to Level 2.

Phase 2: Building the AI Foundation

Once OSS systems are modernised, the next step is to integrate AI and automation.

It involves:

  • Establishing a centralised data platform for real-time analytics.
  • Developing Machine Learning models to detect patterns and optimise performance.
  • Mapping full network topologies to support AIOps (AI-driven operations).

For TM Forum, thisaligns with CSPs aiming for Level 2 to Level 3, where AI-driven predictive capabilities become viable.

Phase 3: Closed-Loop Automation

With AI models in place, the focus shifts to implementing automated and self-regulating processes.

It involves:

  • Domain-level closed-loop automation (e.g., optimising radio access networks).
  • Resource and service-level automation, aggregating key performance indicators (KPIs) for real-time decision-making.

For TM Forum, this corresponds to Level 3 to Level 4, where networks can proactively detect and resolve issues.

Phase 4: Full AI-Driven Autonom

The final step is enabling self-configuring, self-healing and self-optimising capabilities, transforming networks into truly autonomous systems.

It involves:

  • Expanding automation to order management and service experience.
  • Integrating Digital Twins to simulate network behavior and predict failures.
  • Continuously refining AI models through business process reengineering.

For TM Forum, this phase supports CSPs reaching Level 5, where the network functions without requiring human intervention.

The Role of AI and Digital Twins in Enabling Autonomy
AI enables predictive maintenance, real-time traffic optimisation, and automated fault resolution, ultimately forming the backbone of autonomous networks. Gartner emphasises that Digital Twins – virtual replicas of physical networks – are becoming a key enabler of self-governing networks.

As Gartner states, “Transforming OSS into an Autonomous Operations System (AOS) will revolutionise network automation, orchestrating complex processes with minimal human oversight.” This aligns with TM Forum’s vision that higher automation levels require intelligent, AI-driven OSS architectures.

What should CSPs do?

So, to successfully transition towards fully autonomous networks, CSPs should focus on the following strategies:

  1. Develop a structured automation roadmap: align network automation goals with business objectives and customer needs.
  2. Invest in AI-ready OSS platforms: adopt cloud-native, API-driven OSS architectures that support real-time analytics.
  3. Leverage Digital Twins: use network simulation models to improve decision-making and reduce operational risks.
  4. Adopt closed-loop automation: shift from reactive to proactive network management, ensuring better service quality.
  5. Collaborate with technology partners: engage with leading AI and cloud providers to accelerate OSS modernisation and AI adoption.

The journey towards autonomous networks is complex but achievable. By integrating TM Forum’s automation maturity model with Gartner’s phased OSS transformation, CSPs can create a structured and AI-driven roadmap to achieve full autonomy. As AI and Digital Twins continue to evolve, CSPs that strategically modernise their OSS will gain a competitive edge, enabling seamless and intelligent network operations.

At Celfocus, we have been helping clients progress to Level 4 and Level 5 for several years now, integrating AI and automation to enhance customer experience and operational efficiency. However, as networks become increasingly complex, achieving full autonomy requires greater visibility into network resources and new approaches to data representation.

Celfocus approaches digital network transformation by first tackling process and solution design with an E2E holistic perspective covering the entire CSP lifecycle (from planning to assurance). To do this, we focus not only on the network-centric process itself but also on business interactions and their value by using our proprietary facilitation maps technology, bringing all the information together to be discussed by multiple teams. This approach allows us to break existing silos and increase synergies between business, engineering, and operations teams, while also working towards reducing the complexity and overlapping of solutions. With the process and solution design in place, our vision is to achieve a fully digital network experience for customers powered by a digital network management platform that contains all the right technological foundational capabilities to boost teams’ efficiency and TTM while delivering automated processes that return value to both business and operations.

Our approach is inspired and refined from our long-term collaboration with CSPs, including deployments in 12+ countries on a global level. It leverages large amounts of data produced by CSPs, combining advanced analytics, Artificial Intelligence & Machine Learning (AI & ML) and orchestration & automation solutions combined, to enable operators to automate planning, readiness, fulfilment, and assurance with minimum human intervention.

Following TM Forum’s model as well as Gartner’s guidelines, we are embracing a new wave of innovation, collaborating with partners and clients to bring greater intelligence to network operations and management. Our solutions have already proven instrumental in enhancing service impact analysis and root cause identification. Yet, this is just the beginning – Digital Twins unlock new possibilities that will drive even greater value for operators in the future.

Learn more about Celfocus’s Autonomous Networks Approach here.

About the author

Inês Rocha is a Marketing Consultant at Celfocus with over seven years’ experience in content strategy and creation. With a strong focus on Technology, Artificial Intelligence, and Digital Transformation, she specialises in crafting impactful narratives that inform, inspire, and support business impact. She holds a degree in Communication Sciences from NOVA University Lisbon and a postgraduate qualification in Storytelling.

NorthC acquires six Colt data centres in Germany and the Netherlands


The transaction includes data centres in key metropolitan regions including Frankfurt, Berlin, Hamburg, Munich, Düsseldorf and Amsterdam

Colt announced the divestment of six of its data centres in major cities across Europe to regional data centre provider NorthC, headquartered in the Netherlands and majority-owned by funds managed by DWS Group. Colt said it will also divest two of its data centres in London to a UK-based data centre business, also owned by funds which are managed by DWS Group. 

The sale expands NorthC’s European data centre footprint in the Benelux and DACH regions and enables Colt to focus on its core business strategy. This acquisition further increases NorthC’s capabilities to serve customers in multiple regions in the Benelux and DACH and establishes nationwide coverage in Germany. It also adds significant available capacity in Amsterdam, one of NorthC’s core markets.

The divestment is expected to complete later this year. The data centres were part of the assets Colt gained with its acquisition of Lumen EMEA in 2023.  The data centres collectively have a power availability of over 25MW. The colocation business of approximately 400 customers will transfer from Colt as part of the divestment. The majority of these customers also purchase network products from Colt and will remain Colt customers.   

Colt will enter into a long-term partnership agreement with NorthC and will retain network equipment in the divested data centres as part of its global digital infrastructure. Colt will remain an important customer in the to be acquired data centre facilities.

“This represents another major milestone in our journey to operate the leading platform of regional data centers in northwestern Europe,” said NorthC Group CEO Alexandra Schless (above). “Germany, as Europe’s largest economy, is a key strategic market for us. With this acquisition, we will strengthen our presence across key economic regions in Germany, which will unlock further growth and new opportunities.”  

She added: “We’re also pleased to further expand our partnership with Colt as one of the major network providers within the rich connectivity eco-system in each facility.”

Colt CEO Keri Gilder said: “For Colt, the sale enables us to focus on building our strengths in key strategic areas, driving growth, committing to an outstanding customer experience and building sustainable digital infrastructure.”

Colt made that point that the sale by Colt Technology Services was entirely separate and unrelated to Colt Data Centre Services, which has operated as a separate entity to Colt since 2023. Colt Data Centre Services designs, builds and operates data centres for global hyperscalers and large enterprises.     

NorthC has been advised by Evercore, acting as M&A advisor and Latham & Watkins serving as legal counsel on this transaction.

Telco to techco: Cultivating the right culture and approaches to GenAI

Data and AI guru Richard Benjamins argues that intelligent regulation boosts innovative use of GenAI, and explains the pros and cons of three main approaches to leveraging the tech

Until March 2024, Richard Benjamins was Chief Responsible AI Officer at Telefonica and founder of its AI for Society and Environment initiative, where he played a pivotal role in shaping Telefónica’s approach to the ethical use of AI. He is co-founder and CEO of the Spanish observatory for ethical and social impacts of AI (OdiseIA), founder and Co-CEO of RAIght.ai, a responsible AI startup and chair of EIT Digital’s Supervisory Board.

Benjamins was in conversation with Mobile Europe‘s editor, Annie Turner, at our most recent virtual Telco to techco conference. This article only briefly highlights some of the topics we covered – watch the whole session on video now.

We started by discussing GenAI models. Benjamins said he doesn’t think there is much point in telcos creating their own large language models (LLMs) as there are a number of LLMs available around the world that “can be perfectly used by telecommunication operators and any business around the world”.

He thinks a good way to think about LLMs is that they are like the public cloud – “in that most of companies across the world will use those services on the cloud. That doesn’t mean that they can be profitable or or innovative, etc because they will run it on top of that.”

He warned, “It’s very hard for an individual organisation to keep the pace with those big companies that invest a huge amount of money in this technology. If a telco wants to…build such a model by themselves, they need to be aware that they need a huge amount of resources, a huge amount of skills to be better. Otherwise, you have your own language model, but it’s worse in all aspects, security, safeguarding costs, etc. compared to others.”

Three ways to leverage LLMs

Benjamins went into detail about the three main ways for an enterprise to take an existing model and apply its own data/documents to train the model for specific cases. He discussed the potential benefits and risks of each one, including the amount of investment and effort they require. The three approaches are: extended prompting in which APIs play a key role; retrieval augmented generation (RAG); and fine tuning a large model that exists with an enterprise’s own documentation.

He concluded, “Companies have to experiment [to find] which is is better, but shouldn’t forget that extended prompting is the best way to start and to test.”

Small can be good too

He is a fan of small language models for certain use cases, explaining, “a large language model can have from a north of 200 billion parameters: a small language model only has in the order of 4 billion [which] is still [such] a huge amount that we cannot even imagine what it means. But in terms of consumption and in terms of energy, the resources needed and the cost of development, it is much cheaper.”.

Benjamins talked about the growing use of GenAI by individuals at work and highlighted some of the dangers, such as around data privacy. He thinks all use of AI needs awareness and training, and transparency, in writing a document say, that it was largely produced by GenAI. He noted, “In the end you are always accountable as a person for what you send out, even if it’s 100% generated with GenAI.”

Generates info rather than retrieves

He added, “Do a read through of what the system generates, and check whether there are those bias elements, and then, if you detect them, remove them rather than just sending out everything as it comes back. Also, you have to check the facts, because Generative AI, as the word says, generates information rather than retrieves information…oftentimes they are in sync but sometimes that’s where the hallucinations come. It generates things that could be true, but actually are not true. So you also have to check that and take accountability for that.”

Regulation does not squash innovation

Sometimes regulation is seen as the opposite of innovation, and if we have one, then the other is stifled. There has been particularly heated debate about this concerning GenAI. Benjamins said, “The trade-off between regulation being responsible and innovation, I think is largely an overhyped discussion that is used anywhere, at any occasion, without a lot of depth.”

“Let me explain it, he continued. “I think there are many business benefits for being responsible with AI. The first is that investors require it more and more [as part of looking into companies’ Environmental, Social and Governance] policies.”

He acknowledged that in the current geopolitical situation there is strong pushback on any regulation, even concerning climate so, he said, “I think self-regulation in this respect is even becoming more important in the coming years, even though I think regulation will catch up later again.

Next Benjamins argues that employees and customers demand more and more responsible behaviour from companies, and do no only focus on profits, but the impact of the company, in society and the environment. Some studies found that more than “65% of employees look at those things, especially if we think about AI and data. And there is a enormous competition for talent. It is very, very hard to get the right talent.” Going about things responsbility makes it easier to attract and keep talent, according to Benjamins.

Thirdly, a highly persuasive argument from Benhamins is that, “If you have a governance model, you have things in place that detect potential problems early on. There is a system in place that can do that with roles, responsibilities, etc. That means that some people, some companies, start to innovate more and faster because…if you have this safety net, people feel safe…it’s like a sandbox. They can experiment, because they know if they do something they shouldn’t. Somebody will catch it. And that drives innovation in the culture of information, because people feel a secure in a place where they can innovate. So these are all reasons why governance and guardrails in place also gives creative people freedom,

Telecoms in poll position

“I’m very pleased to say that the telecommunications industry is leading in terms of responsible use of AI,” Benjamins states.

He points to The GSMA Responsible AI Maturity Roadmap launched last September, based on input from many operators, including Telefonica, to help operators benchmark their progress and understand the whole picture and offer recommendations about how to progress. For example, the highest level, 5, requires a governance model with roles and functionalities, to ensure all AI systems are registered and analysed for risks.

He adds, “Being responsible can include regulation, or it can be self-regulation, or it can be international recommendations, like OECD, or Unesco or the Council of Europe – there are many international recommendations are out there.”

Mind gyms

In something of an intriguing aside, Benjamins said that as tech has taken the physical effort labour out of so very many daily tasks, we now need to make an effort to keep fit, such as going to the gym. Benjamins thinks we will need “mind gyms” as we rely more and more on GenAI to do certain types of mental heavy lifting…

Watch the whole session on video now.

On-prem cloud: reconciling cloud benefits with data sovereignty

Partner content: Operators face cyberattacks, more regulation, the rise of AI and network APIs, while legacy infrastructure has reached its limits – the cloud conversation is increasingly urgent

Telecom operators face a new reality: the risks of cyberattacks, increased regulatory pressure, the rapid proliferation of AI, and accelerated network API exposure. Their legacy complex infrastructure has reached its limits. This is why the conversation around cloud adoption has become increasingly urgent.

The inevitable shift to cloud hosting

Embracing this transition unlocks numerous benefits for telcos: it enables true innovation while maintaining the control and security they require. Key drivers of this shift include:

  • Simplicity: Infrastructure-agnostic applications streamline feature implementation and reduce dependencies, allowing operators to adapt quickly to market demands without being bogged down by complex legacy systems.
  • Quality assurance: Teams can test binary artifacts across environments, identifying errors early and focusing on value creation rather than troubleshooting infrastructure issues.
  • Availability: 99.99% uptime is essential for mission-critical operations. Cloud platforms are designed to support a high level of reliability and maintain reliability even during upgrades or failures.
  • Velocity: CloudOps and containerized application delivery accelerate feedback loops and feature development pipelines, boosting customer satisfaction.
  • Scalability: The explosion of data and connected devices demands flexible solutions to store and process increasing amounts of network metadata and respond to evolving resource needs of solutions.

The case for on-prem cloud for telcos

For many telecom operators, the idea of moving to the cloud raises valid concerns about data sovereignty and compliance. According to a recent Barclays survey, 83% of enterprises are planning to migrate workloads back from public cloud environments to private cloud solutions, demonstrating a growing need for greater control and security. On-premise cloud solutions offer a compelling middle ground that combines the benefits of cloud computing with the security and control that telcos need.

  • On-premise cloud solutions allow operators to maintain full control over their data while still taking advantage of cloud-native technologies. This ensures compliance with local regulations and addresses concerns about data privacy. Furthermore, these solutions are flexible and can be deployed on existing container infrastructures, virtual machines, or dedicated hardware—allowing telcos to choose the deployment method that best suits their needs.
  • Security is another area where on-premise cloud excels. With robust features such as end-to-end encryption, advanced access controls, and regular security updates, these solutions can enhance an operator’s overall security posture rather than compromise it.
  • Cost efficiency is also a significant advantage of on-premise cloud solutions. By optimizing resource utilization and adopting a pay-as-you-go model, telecom operators can effectively manage costs while scaling their operations as needed.

Navigating the Transition with Intersec

Transitioning to an on-premise cloud environment doesn’t have to be a daunting task. Telecom operators can take several strategic steps to ensure a smooth migration by partnering with vendors who understand the unique requirements of the telecom industry.

Leveraging 20 years of expertise with mobile network operators and a decade of experience serving governments, Intersec has developed an industry-specific cloud offerings designed to facilitate seamless and rapid deployment of new versions—a critical factor in advancing telecom geolocation techniques.

  • Intersecrolls out software in container format instantly using Red Hat OpenShift-certified containers. This allows telecom operators to maintain full control over data privacy and sovereignty while benefiting from the advantages of cloud computing, including optimal performance and reliability, enhanced security and compliance, optimized resource utilization, automation, and orchestration.
  • Additionally, the Intersec cloud approach is entirely agnostic to the underlying infrastructure, supporting multiple deployment options—from straightforward container delivery to a fully managed and hosted solution. Customers can host software in their existing cloud environment or rely on Intersec’s hosting services, which include software upgrades, security fixes, scalability planning, audit logs, contingency plans, and more.
  • Importantly, Intersec adopts a phased CloudOps approach to mitigate risks associated with migrating critical systems to the cloud. This strategy relieves IT teams from labor-intensive maintenance tasks, reduces infrastructure investments, and ensures an immediate return on investment through instant product releases and upgrades.

Conclusion

The future of telecommunications lies in embracing cloud technologies. By adopting on-premise cloud solutions, telecom operators can position themselves at the forefront of innovation while maintaining the control and security necessary for their operations. As we move deeper into an era defined by 5G and beyond, those who embrace this transition will be best equipped to meet emerging challenges and seize new opportunities in an increasingly competitive landscape while maintaining full data sovereignty. 

Live  Q&A session invitation

To learn more, join Intersec and Red Hat for an exclusive live session on Wednesday, May 7 at 14:00 CEST on “Cloud Flexibility, Data Sovereignty”, where industry experts will explore the transformative benefits of Red Hat OpenShift containerized deployments for Telecoms and Governments. Click here to register.

Speakers:

  • Mark Longwell, Director of Telco and Edge Alliances at Red Hat
  • Jean-Marc Coïc, CTO at Intersec
  • Sébastien Synold, Project Manager at Intersec

About the author

Sebastien Synold is a Product Manager at Intersec. He has over 25 years of experience in the networking and telecom industries, with specific expertise in traffic intelligence and network-based geolocation. Sebastien currently oversees the Intersec Agora platform, a cloud-native application that collects telecom metadata and delivers exclusive insights for safety, security and commercial use cases.

Netia offers Huawei’s fibre-to-the room on its gigabit services


The Polish operator is bundling Wi-Fi 7 as well to give full household coverage even in larger premises

Polish operator Netia announced that from 15 April this year, customers who purchase the faster fibre broadband options (up to 1Gbps and up to 2Gbps) on its own network in single-family housing will receive Huawei’s fibre to the room access kits with Wi-Fi 7 at no extra cost. These kits enable full fibre speeds in every room – even in larger homes. Netia said it is the first operator in Poland to offer such a solution.

FTTR, the solution developed by Huawei, typically includes two devices (a router with a built-in ONT and a remote unit), featuring a 2.5GE port and Wi-Fi 7. Together, they create a unified home network similar to mesh systems. Because the remote unit can be connected to the router using fibre optic cable (as well as via Ethernet or Wi-Fi), the home network can be configured to deliver full fibre speeds – up to 2Gbps – in every room.

“Huawei FTTR is currently the most advanced solution for home fibre networks, combining the speed of fibre with the convenience of Wi-Fi 7, in which Netia is a nationwide leader. After success in Asian countries, FTTR has also been very well received by customers of leading operators in several Western European countries,” said Netia head of B2C offer development and converged projects Piotr Rożalski. 

“Netia is the first operator to make this innovative solution available to users in Poland. We are confident it will be appreciated by our demanding customers,” he added. 

The FTTR kit available from Netia for subscribers on the up to 1Gbps and up to 2Gbps plans under the “Nieziemska rozrywka III [domy 200] PON offer” includes the Huawei V261a-20 router with ONT and the Huawei K251a-20 remote unit. The equipment supports Wi-Fi 7 and can aggregate both 2.4GHz and 5GHz bands (the end device can use both bands simultaneously, or seamlessly switch to a single band in case of signal loss without any degradation in performance).

No extra cost

In February this year, Netia – being the first nationwide internet provider in Poland using GPON fibre technology – introduced customer devices with 2.5GE LAN Ethernet ports and Wi-Fi 7 connectivity, which were offered at no additional cost to new subscribers of the up to 2 Gbps tariff on Netia’s network in multi-family housing.

For several years, Netia has offered its subscribers the option to lease various solutions to improve the range of fast, fixed internet access in flats or houses, starting from just PLN 5 per month. The operator said this is particularly useful in houses, where the number of floors, overall volume, and number of rooms make solutions like Mesh especially effective. 

- Advertisement -
DOWNLOAD OUR NEW REPORT

5G Advanced

Will 5G’s second wave deliver value?