5G Non-Terrestrial Networks (NTN) extend cellular connectivity to underserved regions by combining satellite technology with 5G New Radio standards, complementing rather than replacing terrestrial infrastructure. Standardized in 3GPP Release 17, NTN operates through Low Earth Orbit (LEO), Medium Earth Orbit (MEO), or Geostationary Orbit (GEO) satellites using ground-based or onboard gNodeBs. By mid-April 2026, 21 operator-satellite partnerships had deployed direct-to-cell services, with 57 operators across 35 countries offering commercial satellite connectivity. 5G NTN advantages over 4G include Doppler compensation, flexible numerology, and adaptable HARQ operation. Physical constraints include approximately 5 dB signal-to-noise ratio on phone uplinks and roughly 6 Mbit/s aggregate capacity per 5 MHz satellite beam. Standardized mobility procedures enable seamless transitions between satellite and terrestrial networks, though many current services rely on coverage reselection rather than automatic handover. For regions like Nepal with mountainous terrain and remote settlements, NTN offers economically viable connectivity where conventional infrastructure deployment proves impractical.

Direct-to-cell (DtC) satellites operating from low Earth orbit provide LTE connectivity to unmodified phones, but face significant technical challenges related to Doppler frequency shifts. As satellites move at orbital altitudes of 525-530 kilometers, they create Doppler shifts reaching approximately ±40 kHz across a pass—far exceeding the ±7.5 kHz compensation capability of standard 15 kHz subcarrier spacing used in LTE band 25. Although satellites pre-compensate Doppler shifts for a single beam reference point, user equipment elsewhere in the beam must handle residual offsets up to 10 kHz, causing handover failures and radio-link dropouts. This study corroborates these effects through detailed measurements of Starlink transmissions and proposes an improved compensation method. By implementing user equipment cell-specific pre-compensation based on the UE's location relative to the beam reference point and known satellite position, the researchers demonstrate enhanced handover efficiency. The technique enables reliable neighbor cell identification and successful handovers, preventing radio-link failures and substantially improving direct-to-cell service quality for standard LTE devices.

floLIVE, a London-based company with Israeli engineering, has addressed a critical gap in IoT connectivity by building a distributed mobile core network focused on data routing efficiency. While the GSMA's SGP.32 eSIM IoT specification handles profile management through cloud-based eIM systems and SM-DP+ servers, it does not determine where data actually travels after a profile is activated. floLIVE recognized that data from UK devices on UK networks can unnecessarily route through foreign gateways depending on the SIM provider's infrastructure, regardless of SGP.32 compliance. In February 2026, floLIVE achieved full SGP.32 support through partnership with Kigen, demonstrating that standards-compliant eSIMs and intelligent traffic routing are complementary but distinct capabilities. Founded around 2015-2016 and backed by a $4.2 million Series A in February 2017 from ARIE Capital and Golden Sunflower Capital, floLIVE positioned itself as an end-to-end IoT networking provider. The company's core innovation lies in optimizing the data path itself, making it essential for organizations seeking both eSIM flexibility and efficient packet routing in IoT deployments.

This paper addresses a critical challenge in sixth-generation (6G) wireless networks by proposing a hierarchical multi-agent framework for Intent-Based Networking (IBN) that translates natural language operational intents into executable network configurations. The system employs Large Language Model (LLM) based agents operating through iterative reasoning-acting (ReAct) cycles, featuring an orchestrator agent coordinating specialized Radio Access Network (RAN) and Core Network agents. Unlike existing approaches that rely on inflexible rule-based systems or opaque end-to-end neural models, this architecture balances interpretability with operational constraint enforcement. Experimental evaluation demonstrates superior performance compared to monolithic agents, rule-based systems, and direct LLM prompting across diverse benchmarks. The findings reveal that while contemporary LLMs possess substantial telecommunications knowledge, structured prompt engineering incorporating domain-specific decision thresholds and operational policies is essential for reliable network automation. The proposed architectural principles are directly applicable to Open RAN (O-RAN) deployments, offering a practical pathway toward zero-touch network management envisioned for 6G systems.

# Summary The integration of satellite technology into 5G networks represents a significant evolution in global telecommunications infrastructure. This article explores how satellite systems are being incorporated into next-generation mobile networks to extend coverage beyond terrestrial limitations. The piece discusses the technical frameworks enabling this convergence, including non-terrestrial networks (NTNs) and satellite-to-5G integration standards being developed by telecommunications bodies. Key applications include bridging connectivity gaps in remote regions, supporting IoT deployments, and enhancing network resilience during emergencies. The article contextualizes this within broader digital sovereignty initiatives, particularly Europe's Iris2 constellation program, which aims to provide autonomous satellite-based connectivity independent of external dependencies. This integration matters for global telecommunications equity, as it enables ubiquitous coverage while addressing infrastructure gaps in underserved areas. The convergence of satellite and terrestrial 5G networks is reshaping how data centers and cloud services can operate globally, creating new commercial opportunities in the space technology sector while supporting international digital infrastructure independence.

Spain's telecommunications sector has long benefited from a stable regulatory framework that balances investment incentives with competitive market dynamics through cross-party consensus. However, this foundation faces potential disruption through a proposed deregulation of Telefónica's fiber duct access—a critical physical infrastructure supporting over eighty percent of alternative fiber deployments across the country. The proposal would replace existing regulations with unilateral commitments by Telefónica and significantly revise access pricing, with potential increases reaching 116 percent over five years according to regulator documentation. The debate extends beyond price implications to question whether market conditions justify abandoning a proven regulatory model. Fiber duct infrastructure represents a natural monopoly with inherent characteristics that make replication economically unfeasible for competitors. The existing MARCo service, which establishes cost-oriented wholesale access pricing for ducts and conduits, emerged from fifteen years of practical deployment experience rather than abstract regulation. Previous attempts by cable and ADSL operators to build alternative infrastructure demonstrated substantial economic and operational challenges, establishing the current framework's necessity for ensuring competitive parity in retail markets.

Nokia and Nvidia have announced significant progress toward commercially deploying AI-RAN technology, which leverages accelerated computing to enhance radio signal processing and enable artificial intelligence applications at the network edge. T-Mobile successfully conducted the first AI-RAN call on its commercial network in Bellevue, Washington, while European operators including BT, Elisa, Deutsche Telekom, and Orange are conducting trials and validation efforts, with BT planning Birmingham trials in 2027. The technology represents a fundamental shift in telecom infrastructure by enabling computing within network operators' systems rather than relying solely on distant data centers, thereby reducing latency for AI applications. Deutsche Telekom's laboratory trials demonstrated 10 to 15 percent improvements in uplink speeds through optimized signal processing. Nokia and Nvidia are also developing a sovereign agentic AI platform to automate network management while maintaining human oversight. While the initiative offers operators improved equipment utilization and new revenue opportunities, it demands substantial capital investment in an industry already constrained by tight spending discipline. Nokia contends that AI-RAN can enhance existing infrastructure profitability and justify new investments aligned with business growth.

SGP.32 is the GSMA standard for remote SIM provisioning in IoT devices, enabling organizations to download, switch, and manage eSIM profiles over the air without requiring physical device interaction. Unlike earlier standards such as SGP.02 for M2M and SGP.22 for consumer devices, SGP.32 features lightweight protocols specifically designed for IoT scalability and low-power performance, eliminating complex provisioning processes. The standard supports centralized fleet management across thousands of devices globally while minimizing power consumption for battery-powered equipment. SGP.32 incorporates enhanced encryption and authentication mechanisms for secure provisioning and profile management, with full compatibility across LTE-M, NB-IoT, 4G, 5G, and satellite networks. Com4 now offers SGP.32-ready IoT SIM and eSIM solutions, leveraging Nordic engineering and private core infrastructure to deliver secure, automated, and globally scalable IoT connectivity. This advancement significantly simplifies large-scale IoT deployment management throughout device lifecycles.

Embedded SIM technology (eSIM) combined with embedded Universal Integrated Circuit Card (eUICC) software is revolutionizing IoT device deployments by enabling remote provisioning without physical access. Unlike traditional removable SIM cards, eSIMs are embedded directly into devices, allowing network operators to send SIM profiles over-the-air (OTA) after deployment. This approach delivers significant advantages for IoT implementations, including enhanced scalability, streamlined supply chain operations, and support for low-power connectivity standards like LTE-M and NB-IoT. The ability to provision devices post-deployment proves particularly valuable for applications like connected vehicles, where final destination is unpredictable at manufacturing. Organizations can now deploy identical hardware globally and configure network connectivity based on actual operating location, eliminating costly SIM card replacement logistics. The GSMA SGP.32 specification, most recently updated to version 1.2 in June 2024, establishes standards for remotely provisioning and managing eUICCs in IoT devices with constrained networks or no user interface. This standardization framework enables manufacturers and IoT providers to implement eSIM technology reliably across diverse deployment scenarios.

SGP.32, the GSMA's latest eSIM specification, represents a transformative advancement for Internet of Things deployments, specifically addressing the limitations of previous remote SIM provisioning standards. Introduced to overcome constraints associated with physical SIM logistics and operator lock-in, SGP.32 unifies the strengths of consumer-focused (SGP.22) and legacy M2M (SGP.02) frameworks into a single, purpose-built standard designed for autonomous, resource-constrained IoT devices. The specification enables zero-touch connectivity management through direct eUICC integration, allowing enterprises to embed Universal Integrated Circuit Cards directly onto devices in any form factor. A critical innovation is SGP.32's capacity for mass remote profile switching, permitting organisations to seamlessly transition network operator profiles over-the-air across tens of thousands of distributed units simultaneously. The standard also introduces bi-directional triggering capabilities, supporting both server-driven and device-initiated commands, unlike predecessor standards. For enterprises managing globally distributed IoT fleets—from smart utility meters to connected vehicles and logistics trackers—SGP.32 fundamentally streamlines international operations by eliminating physical interaction requirements and reducing operational overhead associated with traditional deployment methodologies.

We have spent decades building networks primarily for humans. Voice, messaging, internet access, and eventually video. The objective was to transport information as efficiently as possible. Now synthetic intelligence is becoming another customer. Robots, autonomous vehicles, drones, industrial machines, and software agents capable of interpreting information and taking actions. Please don’t confuse these systems with their rather dumb IoT ancestors. Intelligence will live in different places. Small models in devices, more powerful models at the edge, and massive models in data centers. They will use different processors, consume different amounts of power, and operate under different latency requirements.

On November 4, Qualcomm announced that SpaceX had selected its Dragonwing platform to support tens of thousands of terrestrial 5G macro sites, with deployment beginning in 2027. The companies highlighted 64T64R Massive MIMO at 2 GHz, advanced radio processing, and a future path toward AI-native 6G. It sounds impressive, particularly if you have never built a mobile network. But I see so many problems that they go well beyond chipset choice. I 100% believe this partnership is not the end game for Musk, and he will keep iterating (or making mistakes) until he gets it right.

Telcos have historically sucked at building ecosystems. We tried app stores, social platforms, content, advertising, and developer platforms. We had millions of customers, billing relationships, distribution, and money. Yet, with a few important exceptions, we failed to turn those advantages into businesses that attracted developers, creators, and partners at scale. We keep announcing ecosystems. Other companies keep building the ones people actually use. That history should worry us as we approach 6G.

We are moving toward a world where communications, sensing, and intelligence increasingly share computing infrastructure. And for the first time, the software used to develop advanced radio algorithms could belong to the same computing ecosystem used by millions of developers building robots, autonomous vehicles, medical applications, and AI systems. I think we still get five things wrong about GPUs, especially when discussing AI-RAN and 6G.

On November 3, Perry, Ohio, residents voted overwhelmingly to demand financial guarantees from data center developers seeking tax incentives. Some 87% supported the measure. In nearby Pataskala, 74% voted to prohibit large data centers. These are preliminary election results, but they send a pretty clear message. And this goes well beyond a few unhappy communities. According to the University of Pennsylvania’s Annenberg Public Policy Center, 61% of Americans oppose new data centers in their neighborhoods, up from 49% just four months earlier. Pew Research finds that 52% of Americans are more concerned than excited about AI, compared with 37% in 2021. Another 71% expect AI to eliminate jobs. Meanwhile, Gartner reports that 65% of consumers believe brands already produce too much AI-generated content. The FBI recorded almost $900 million in reported losses from AI-related fraud complaints in 2025. I find these numbers fascinating because we are seeing two completely different reactions to the same technology. Silicon Valley celebrates another extraordinary technological revolution. A growing part of the population sees job insecurity, higher electricity demand, unwanted infrastructure, and an internet increasingly filled with synthetic garbage.

SpaceX is preparing to spend tens of billions building a terrestrial mobile network. But I suspect Musk is looking for something much bigger than selling smartphone subscriptions. Here are five possible explanations, and how well they hold up economically.

AI is creating enormous new businesses, but telecom operators are barely participating. While the industry concentrates on efficiency, networks, and GPU infrastructure, other companies are moving into markets that telcos should be fighting for.

Machine learning is increasingly transforming optical access networks, which form the critical infrastructure connecting service providers to end users. This article examines how artificial intelligence and machine learning algorithms are being applied to optimize network performance, enhance resource allocation, and improve operational efficiency in optical communications systems. Key applications include predictive maintenance, signal quality optimization, and automated network management—technologies that enable real-time monitoring and adaptive responses to network conditions. These techniques leverage data analytics to identify patterns in network behavior, predict potential failures before they occur, and automatically adjust optical signal parameters to maintain service quality. The significance of this advancement lies in addressing growing bandwidth demands while reducing operational costs and minimizing service disruptions. By implementing machine learning in optical access networks, telecommunications providers can achieve superior network reliability, faster problem resolution, and more efficient spectrum utilization. This integration represents a major shift toward intelligent, self-optimizing network infrastructure that adapts dynamically to changing demands, ultimately delivering improved customer experiences and competitive advantages in increasingly data-intensive telecommunications environments.

The European Space Agency, European Commission, and over 20 European space companies have launched a collaborative initiative to develop Eagle-1, Europe's first space-based quantum key distribution satellite system. This partnership, formalized through a contract between ESA and Luxembourg-based SES, aims to establish an unbreakable, physics-based secure communications network for the European Union. Eagle-1 will demonstrate quantum key distribution technology, which leverages quantum mechanics to distribute encryption keys while immediately detecting any eavesdropping attempts. The satellite will build upon technologies developed under ESA's Scylight programme, validating critical EU-sourced components. The system will operate from low Earth orbit to ground stations, including upgraded optical terminals from Germany's DLR and a new terminal from the Netherlands, with Italy's Sitael providing the platform and Germany's Tesat Spacecom supplying the quantum payload. Scheduled for launch in 2024, Eagle-1 will undergo three years of in-orbit validation, ultimately supporting the European Quantum Communication Infrastructure and enabling the EU to achieve sovereign, autonomous cross-border quantum-secure communications that significantly enhance European cybersecurity autonomy.

5G non-terrestrial networks (5G NTN or NR-NTN) represent a fundamental evolution in mobile connectivity, leveraging satellites across low Earth (LEO), medium Earth (MEO), and geostationary (GEO) orbits, alongside high-altitude platforms (HAPS), to extend coverage to oceans, remote regions, and disaster zones. Standardized beginning with 3GPP Release 17, with subsequent enhancements in Releases 18 and 19, 5G NTN has transitioned from research and prototypes into practical deployments and early service rollouts. The technology operates across FR1 (410 MHz to 7.125 GHz) and FR2 (24.25 GHz to 71 GHz) spectrum bands, with recent expansion into Ka and Ku bands and future targeting of Upper C-band, Q/V, E, and D bands. The architecture comprises four segments—ground, air, space, and network—supporting diverse deployment models. Two satellite approaches exist: simple transparent relays enabling faster deployment but with higher round-trip times, and regenerative payloads offering onboard processing and reduced latency at increased complexity. Managing interference across wide coverage areas requires specific metrics including power flux density (PFD) and equivalent power flux density (EPFD). Physical layer adaptations address Doppler effects, propagation delays, and pathloss through advanced antenna systems and beamforming, with comprehensive protocol stack modifications required across all network layers.