Future network technology: Speed alone is not enough

IT in practice9 Oct 202613 min read

Network technology illustration with an Ethernet switch, fibre and wireless networking, a laptop and a connected digital globe. — Illustration with German lettering, not a technical wiring diagram or product photograph. TN, 9 October 2026.

Performance must reach everyday work

When people talk about the networks of the future, the headlines usually feature big numbers. More gigabits, a new Wi-Fi generation and the next mobile technology are supposed to make everything faster.

My first questions are different: does the performance reach the places that need it? Does a connection remain stable when many devices work at once? And can the environment still be managed reliably as more applications are added?

A network should do more than produce a good speed test. It must suit the people, devices and tasks that depend on it.

It is worth looking closely at technical developments. Some are already available. Others are still being standardised. Both matter, but they should not be confused.

Cables still matter, especially for modern wireless networks

More capable devices also raise questions about connections within the building.

Ethernet at 2.5 or 5 gigabits per second is not a future promise. IEEE 802.3bz was approved on 22 September 2016. Completed IEEE standards also exist for 10- and 25-gigabit Ethernet. [2]

That does not mean every workstation should now be upgraded to the highest available speed.

I would plan differently for a computer mainly used for email and straightforward office work than for a workstation handling large video files or extensive technical projects. Central links deserve particular attention: between switches, servers, storage systems and heavily used wireless access points.

For a new installation, I would also consider suitable cable routes, spare capacity and the potential role of fibre connections. Distances, existing cabling and expected demand should determine the choice, not just the speed printed on the box.

A powerful wireless access point cannot compensate indefinitely for an inadequate connection to the rest of the network. The complete path matters here too.

Wi-Fi 7: more than a higher number on the router

The Wi-Fi Alliance launched Wi-Fi CERTIFIED 7 on 8 January 2024. Its features include wider channels and Multi-Link Operation, or MLO. Compatible devices can use several wireless links together to improve throughput, responsiveness and reliability. [3]

The potential peak speed is not the only interesting aspect. Having several usable links can help devices cope with changing radio conditions.

However, 'Wi-Fi 7' does not guarantee that every device supports every feature. Intel's Wi-Fi 7 components, for example, differ in channel width and maximum data rate. Particularly wide 320 MHz channels require compatible equipment and permitted use of the relevant frequency band. [3]

When buying equipment, I would therefore look beyond the router or access point. The laptops and smartphones in use, their software and the access point's wired connection also need checking.

Before a widespread upgrade, I would test the actual workspaces. A meeting room behind several walls tells us more than a measurement right beside the router.

Wi-Fi 8: reliability takes a more prominent role

While Wi-Fi 7 can already be deployed, work on the next generation continues.

IEEE project 802.11bn, the basis for Wi-Fi 8, focuses on Ultra High Reliability. Its project description covers improvements in throughput under particular radio conditions, latency and packet loss, including transitions between wireless cells. As checked on 9 October 2026, the published timeline places major final approvals in 2028. This is a plan, not a completed standard. [4]

I consider that direction sensible. A connection that survives moving between rooms can be more valuable in everyday work than a higher peak rate under ideal conditions.

Nevertheless, I would not postpone a necessary improvement today simply because another wireless generation has been announced. What matters is whether existing technology meets the requirements and which specific shortcoming needs addressing.

Being prepared for future possibilities does not mean waiting indefinitely for working solutions.

Private 5G networks are possible; 6G remains in development

Mobile networks require the same distinction between available technology and future standards.

In Germany, frequencies in the 3.7–3.8 GHz range can be requested for local campus networks. The Bundesnetzagentur, Germany's federal network regulator, publishes the application documents and conditions. [5]

A private mobile network could be an option for a business site with mobile equipment. But I would judge the effort against a concrete task: which devices need connecting, where do they move, what interruptions or delays are acceptable, and who will operate and maintain the additional infrastructure?

For me, a campus network would not automatically replace Wi-Fi or cables. It would be another option that must demonstrate its value.

What has actually been established for 6G

On 17 March 2026, the International Telecommunication Union (ITU) reported draft technical performance requirements for IMT-2030, the forthcoming 6G generation. Draft guidelines for evaluating candidate radio technologies were also completed in June 2026. As checked on 9 October, further formal approvals were still planned. Candidate technology proposals are expected between February 2027 and February 2029. [6]

These are important development milestones. They are not a promise that usable 6G networks will be available everywhere on a particular date.

For current planning, I would therefore monitor 6G as a development rather than treat it as a guaranteed feature of the next installation. The ITU itself explains that technical target requirements do not guarantee real-world deployment performance. [6]

Edge computing: process data closer to its source

Not every task needs to be processed in a distant data centre.

Edge computing brings computing resources closer to devices, users or local networks. In its Multi-access Edge Computing work, ETSI describes deployment options on business premises and at the network edge. Applications include video analytics, connected devices and local content distribution. [7]

One possible example is analysing sensor data locally. Instead of continuously sending all raw data to an external service, an on-site system could decide which results need to be shared.

I would find that particularly useful where short response paths or limited external data transfers matter.

The key question, however, is which dependencies remain. An application is not automatically independent of the internet because some processing happens locally. I would check whether sign-in, licence verification or other required services still need an external connection.

Local processing should provide a clear advantage, not simply add another computer that nobody maintains.

IPv6 belongs in the plan, not on a later to-do list

IPv6 is a version of the Internet Protocol, not a new wireless technology. Its core specification, RFC 8200, was published in July 2017. It provides a much larger address space than IPv4, among other changes. [8]

For new infrastructure, I would not treat IPv6 as merely a tick on a specification sheet. Addressing, name resolution, firewall rules and logging must work together in operation.

This is especially important when IPv4 and IPv6 run side by side. RFC 9099, published in August 2021, discusses operational security for IPv6 networks. A secured IPv4 configuration does not automatically secure every IPv6 connection. [8]

I would therefore require a clear decision: which protocols will be used, how will they be managed, and how will we check that the intended protections actually work?

An unknown second communication path is not future readiness; it is an unresolved operational question.

Security must follow access

A network can be powerful while still granting overly broad access.

In Zero Trust Architecture, published in August 2020, NIST explains that accounts and devices should not be trusted simply because they are inside the local network. Identity, authorisation and access to the particular resource are central. [9]

I would not interpret that as a requirement to replace every existing system with a product labelled 'Zero Trust'.

I would ask concrete questions: why can this device access that service? Does an external colleague need the entire network or just one application? Who removes a permission when it is no longer needed?

A guest device or connected camera should not have the same opportunities as a managed workstation merely because they share a building.

The aim would be an environment where necessary work remains possible without connecting everything indiscriminately.

Encryption must remain adaptable too

Long-term planning also includes whether deployed security methods can be updated.

On 13 August 2024, NIST published its first three final post-quantum cryptography standards. They address establishing protected communications and digital signatures using methods designed to resist attacks by future sufficiently powerful quantum computers. [10]

This is not a reason to claim that every current connection is already insecure. It is a reason to ask about support for new methods in long-lived systems.

For procurement, I would therefore consider update support and a credible migration route. A blanket advertising claim such as 'quantum-proof' is not equivalent to a technical assessment.

AI in networks: understand first, intervene second

Network management is also becoming more automated.

ETSI's work includes Experiential Networked Intelligence and Zero-touch Network and Service Management. Their descriptions cover AI support, automated management and progress towards increasingly autonomous operations. This is standardisation and development work, not proof that every product marketed in these terms already fulfils those goals. [11]

I see a useful starting point where a system helps investigation: combining measurements, highlighting unusual changes and explaining possible causes clearly.

A different level of responsibility arises when it changes access rules or production configurations on its own.

Before allowing those interventions, I would need clear boundaries. Which changes are permitted? How are they documented? How can an error be reversed? And will administrative access remain available if the automation itself fails?

The ability to explain a decision would matter more to me than a particularly convincing recommendation. I want to understand which measurements led to the assessment.

An automatically managed network still needs someone who understands its rules and effects.

What I would prepare today

Not every environment needs the same equipment. These examples show how priorities can differ:

Small office or home office: investigate actual Wi-Fi coverage, reliable internet access, updates and secure remote access first.

Large project files or local AI applications: examine links between workstations, servers and storage, alongside their actual utilisation.

Workshop, warehouse or larger business site: check radio coverage at the work locations, moving devices and, where appropriate, the suitability of a campus network.

Multiple sites: examine application dependencies, alternative connections and consistently managed access rights.

These examples guide an initial assessment; they are not a universal shopping list.

For a new environment, I would also allow capacity for expansion and document the configuration clearly. A major change should include a limited practical trial using the actual devices and tasks.

Ongoing costs matter too. Electricity, licences, maintenance and the ability to replace components later can be more important than a low initial price.

I would not replace working equipment just because another generation has arrived. Equally, I would not tolerate a known problem indefinitely because the next technology is supposedly just around the corner.

My conclusion: future readiness means being able to manage change

Peak data rate alone does not convince me that a network is good.

I want to know whether it works under load, whether access is limited and whether a fault can be narrowed down. New applications should be possible without rebuilding the whole environment every time.

Wi-Fi 7, the development of Wi-Fi 8 and 6G, local data processing and automated management create opportunities. Their value emerges in actual use.

I would therefore neither wait for future technology as a matter of principle nor buy it immediately as a matter of principle. I would judge it by the problem it solves and the additional dependencies it introduces.

The network of the future does not need the latest technology everywhere. It needs a design that leaves room for worthwhile new technology later.

Scope and research

Research date: 9 October 2026. Development targets and announced dates do not guarantee particular product availability or real-world performance. These examples do not replace individual network and security planning.

Sources and further reading

Sources checked on 9 October 2026. Publication dates and planned milestones are distinct from this access date. Linked information comes from the relevant standards bodies, authorities and manufacturers; the Wi-Fi Alliance release was published through GlobeNewswire. Some sources are in German.

[1] IETF: Framework for TCP Throughput Testing, RFC 6349 (August 2011)
Read source ↗

[2] IEEE: Ethernet standards for 2.5, 5, 10 and 25 gigabits per second
Read source ↗

[3] Wi-Fi Alliance: Wi-Fi CERTIFIED 7 (8 January 2024); Intel component specifications
Read source ↗

[4] IEEE 802.11bn project timeline; Qualcomm Wi-Fi 8 explanation (July 2025)
Read source ↗

[5] Bundesnetzagentur: Regional and local networks, including 3.7–3.8 GHz campus networks
Read source ↗

[6] ITU: IMT-2030 / 6G requirements and development milestones
Read source ↗

[7] ETSI: Multi-access Edge Computing
Read source ↗

[8] IETF: IPv6 specification RFC 8200 (July 2017) and operational security RFC 9099 (August 2021)
Read source ↗

[9] NIST: Zero Trust Architecture, SP 800-207 (August 2020)
Read source ↗

[10] NIST: First three final post-quantum cryptography standards (13 August 2024)
Read source ↗

[11] ETSI: Artificial intelligence and automated network management
Read source ↗

Written by TNAll articles