G. EMERGING TRENDS
1. 6G: Beyond 5G
Whilst 5G is still being rolled out in many countries, research and development into 6G is already well advanced. In 2023, the ITU (International Telecommunication Union) formalised the IMT-2030 vision, which sets out the performance targets for 6G for a commercial roll-out envisaged around 2030 (IMT targets–2030 not yet achieved in production). These targets are highly ambitious: peak data rate of one terabit per second (50 times that of 5G), latency of less than 0.1 ms (ten times lower than 5G), connection density of 10 million devices per km² and energy efficiency 100 times greater than 5G.
The technologies being considered to achieve this level of performance are exploring several avenues in parallel. Terahertz bands (THz, between 100 GHz and 10 THz) offer enormous bandwidths for data rates in the order of Tbps, but suffer from extreme atmospheric attenuation, which limits their range to a few metres or tens of metres at best. Their use will likely be limited to confined environments, such as inside buildings, industrial halls and data centres, where short range is not an obstacle.
RIS (Reconfigurable Intelligent Surfaces) are one of the most innovative features of 6G. These are surfaces composed of thousands of passive or semi-active reflective elements, integrated into walls, ceilings or building facades, capable of reflecting and directing radio waves towards users with millimetre precision. They transform the passive environment into a programmable amplifier and reflector, enabling coverage of dead zones or improved signal quality without deploying additional active equipment.
ISAC (Integrated Sensing and Communications) is another key feature of 6G. Rather than using separate systems for communications and sensing (radar, lidar, positioning systems), 6G will use the same radio signals both to transmit data and to ‘sense’ the environment: detecting objects, estimating speeds and trajectories, and mapping spaces. This convergence offers immense possibilities for autonomous vehicles, smart cities and augmented reality.
|
Parameter |
5G (3GPP target) |
6G (IMT-2030 target) |
|
Peak flow rate |
20 Gbps |
1 Tbps |
|
User experience performance |
One hundred megabits per second |
1 Gbps |
|
User-level latency |
Less than 1 ms |
Less than 0.1 ms |
|
Connection density |
1 million devices per square kilometre |
10 million devices per square kilometre |
|
Reliability |
99,9999 % |
99,99999 % |
|
Energy efficiency |
×10 vs 4G |
×100 vs 5G |
|
Location |
10 cm (inside) |
<1 cm (sub-centimetre) |
ℹ 6G R&D programmes
ℹ 6G R&D programmes The European Union is funding the Hexa-X and Hexa-X-II projects (Horizon Europe programme, 25 industrial and academic partners). South Korea is aiming for a commercial 6G launch in 2028 to coincide with the Winter Olympics. Japan is targeting 2030 in its Beyond 5G/6G Promotion Strategy. China is very active in filing 6G patents through its IMT-2030 Promotion Group. Global commercialisation is expected around 2030–2032..
2. Wi-Fi 7 and the evolution of Wi-Fi
The Wi-Fi 7 standard (IEEE 802.11be, EHT, Extremely High Throughput) was finalised in 2024 and the first certified devices are now available on the market. It offers substantial improvements over Wi-Fi 6E, meeting the growing bandwidth demands of virtual reality, 8K streaming and high-quality remote desktop applications.
The key innovation of Wi-Fi 7 is MLO (Multi-Link Operation), which allows a device to use multiple frequency bands—2.4 GHz, 5 GHz and 6 GHz—simultaneously. Up until Wi-Fi 6E, a device could only use one band at a time and had to switch between bands. With MLO, data can be sent simultaneously over multiple links, which increases the aggregate throughput and improves latency by automatically avoiding interference: if the 5 GHz band is congested, traffic is automatically shifted to the 6 GHz band, without any noticeable interruption.
Wi-Fi 7 also introduces 320 MHz channels in the 6 GHz band, doubling the channel bandwidth compared to Wi-Fi 6E (160 MHz). Combined with 4096-QAM modulation (4K-QAM), which encodes more bits per symbol at the cost of increased sensitivity to noise, the maximum theoretical data rate reaches 46 Gbps across 16 spatial streams. In practice, real-world deployments achieve several Gbps under favourable conditions, which is more than sufficient for all current and foreseeable future uses.
ℹ Information
In 2025–2026, Wi-Fi 7 deployments will remain largely limited to pilot sites and high-end equipment.
The next Wi-Fi 8 (IEEE 802.11bn) standard is already being developed by the IEEE working group, with completion expected around 2027–2028. The stated objectives include data rates exceeding 100 Gbps through the use of even wider frequency bands and innovations at the physical layer. Wi-Fi 8 is also expected to enhance multi-access point coordination (multi-AP coordination) to deliver a seamless experience in dense deployments.
3. Green Networking: Networks and the Energy Challenge
The energy consumption of telecommunications networks has become a major environmental and economic issue. Networks and data centres as a whole now account for around 2–3% of global electricity consumption, and this share is growing with the roll-out of 5G, the proliferation of IoT devices, and the explosion in video traffic and AI. Training a large language model such as GPT-4 consumes the equivalent of the annual electricity consumption of several hundred households. Given these realities, energy efficiency has become an essential design criterion for modern networks.
In 5G radio networks, next-generation equipment incorporates sophisticated mechanisms for dynamically putting antennas into sleep mode: when no traffic is detected on a cell for a few milliseconds, the radio transmitters can be put into partial or full sleep mode, reducing power consumption by 60 to 80% outside peak hours. Nokia, Ericsson and Huawei have published ambitious roadmaps to reduce power consumption per transmitted bit by 50 to 80% by 2030, by combining these software optimisations with advances in semiconductor technologies.
In data centres, cooling accounts for a significant proportion of total energy consumption.
Hyperscalers are adopting innovative solutions to reduce this cost
· Direct liquid cooling (DLC), in which coolants circulate directly in contact with the hottest processors, particularly the GPUs used for training AI models
· Immersion cooling, where entire servers are submerged in non-conductive dielectric liquids
· Heat recovery to supply neighbouring district heating networks, transforming heat from servers into a useful resource rather than waste.
Google, Microsoft and Amazon have all made public commitments to carbon neutrality or carbon offsetting for their operations, through Power Purchase Agreements (PPA) with renewable energy producers. Google claims it will power its data centres with 100% real-time carbon-free energy by 2030. Software-based management of network equipment consumption, such as Cisco EnergyWise and Juniper Energy Framework, enables consumption to be dynamically adapted to actual load by disabling unused ports or reducing the clock speed of routing processors during off-peak periods.
4. 5G private networks
Private 5G networks (1>also known as 5G campus networks1>) enable businesses and organisations to deploy their own 5G infrastructure on their premises, without relying on a public operator. This approach offers guarantees of service quality, security and privacy that shared public 5G networks cannot provide, as well as complete independence from incidents on the public network.
In the standalone deployment model, the company deploys the entire 5G infrastructure—base stations (gNB) and the 5G SA core network (5G Standalone Core)—on its own premises. The company’s data never leaves the site. The bandwidth is entirely dedicated to the company’s use. Latency is less than 5 ms between on-site equipment. This model is particularly suitable for factories with automated production lines, ports and airports with intensive logistical requirements, mines and quarries in areas without public coverage, and hospitals with strict medical data confidentiality requirements.
In Europe, Germany has been a pioneer, having allocated a dedicated frequency band (3.7–3.8 GHz) to private industrial 5G networks as early as 2019, enabling companies such as Volkswagen, BMW and Siemens to roll out their own 5G networks in their factories. France is taking a more gradual approach to this issue, with ARCEP having launched consultations on the allocation of local frequencies for industrial users in 2024, with the first allocations expected in 2025–2026.
⚡ Wi-Fi 6E and private 5G convergence
In a modern private network, Wi-Fi 6E and private 5G are not in competition but are complementary. Wi-Fi 6E covers office spaces, fixed warehouse areas and meeting rooms at a lower cost. Private 5G covers outdoor areas, mobile equipment on the move (AGVs (Automated Guided Vehicles)), and applications requiring guaranteed latency and availability that Wi-Fi cannot provide in a disrupted industrial environment.
5. SASE, Network and Security Convergence
SASE (Secure Access Service Edge), a concept formalised by the research firm Gartner in 2019, has become the standard architecture for distributed enterprise networks by 2025. Its emergence responds to a profound transformation in the network landscape of organisations: users work from anywhere (office, home, on the move), applications are in the cloud (SaaS, IaaS), and data traverses networks that are no longer under our control. The traditional model, where all traffic converges on the central data centre hosting the firewall and proxies, has become unsuitable, as it creates bottlenecks and degrades the user experience.
SASE brings together two families of features that were historically separate into a single unified cloud service.
On the network side, SD-WAN intelligently manages multiple WAN links (MPLS, broadband Internet, 4G/5G), selecting the most efficient link for each data stream in real time based on its type (voice, video, critical data)
When it comes to security, ZTNA (Zero Trust Network Access) replaces the traditional VPN by applying the principle of “never trust, always verify”: rather than granting full network access after authentication, each access request to each application is verified individually based on the user’s identity, the security posture of their device and the context of the connection.
The other components of the SASE complement this protection:
· The CASB (Cloud Access Security Broker) provides visibility and control over the use of SaaS applications (file sharing, unauthorised applications)
· The SWG (Secure Web Gateway) filters outbound web traffic and protects against malware
· FWaaS (Firewall as a Service) delivers NGFW firewall capabilities from the cloud, without the need for physical equipment in each branch office
· DLP (Data Loss Prevention) monitors data in transit to prevent leaks of sensitive information.
In 2025, the four leading players in the SASE market according to Gartner’s Magic Quadrant are Palo Alto Networks (Prisma SASE), Zscaler, Cisco (combining Umbrella and Meraki SD-WAN) and Netskope. Cato Networks stands out for its 100% cloud-native architecture, built from the ground up around SASE without any legacy software layers. The current evolution sees SASE converging with IAM (Identity and Access Management) and XDR (Extended Detection and Response) to form SSE (Security Service Edge), which treats security as a continuous layer around users and data, regardless of their location.