Voice Networks and Data Networks: From Analog Telephone Circuits to Modern IP Communication

From AI tools as IS.

For much of the history of telecommunications, a voice network and a data network were genuinely different kinds of networks.

Telephone networks were built primarily to carry human conversations. Computer networks were built primarily to exchange data between machines.

Today, that distinction has become much less clear.

A modern telephone conversation may be converted into digital data, divided into packets, carried over fiber and IP networks, processed by software, transmitted over 4G or 5G radio, and finally converted back into sound at the receiving telephone.

Yet traditional telephone infrastructure has not disappeared everywhere. As of September 2026, legacy circuit-switched, copper-based, and other specialized voice systems continue to operate in several parts of the world while telecommunications companies migrate customers toward digital and IP-based services.

Understanding what changed requires looking at what voice networks and data networks actually mean.

---

What Is a Voice Network?

A useful definition is:

A voice network is a communication network designed primarily to provide voice communication services. Traditional telephone voice networks were generally circuit-switched networks specifically engineered for telephone communication.

Historically, telephone networks primarily carried human speech.

When someone spoke into an old telephone, the microphone converted changes in air pressure into a continuously varying electrical signal.

The conceptual process was:

Human Voice

↓

Continuous sound wave

↓

Telephone microphone

↓

Continuous analog electrical signal

↓

Telephone network

↓

Receiving telephone

↓

Sound

This is why traditional telephone networks are often described as analog voice networks.

However, this needs an important qualification.

Telephone networks did not remain completely analog.

They became substantially digital long before the Internet became dominant.

---

What Was a Traditional Telephone Signal?

Human speech is naturally analog.

Air pressure varies continuously as we speak.

An early telephone microphone converted these continuous variations into a corresponding continuously changing electrical signal.

Therefore:

Sound pressure

→ continuous analog electrical signal

The receiving telephone performed approximately the reverse operation:

Analog electrical signal

→ speaker movement

→ sound pressure

So it is reasonable to say:

Early telephone voice was represented primarily by continuous analog electrical signals, and traditional voice networks were designed primarily to transport those telephone voice signals.

It is more accurate to say voice signal rather than simply telephone signal, because a telephone network also carries signaling information associated with dialing, ringing, call establishment, supervision, and disconnection.

---

Traditional Telephone Networks Used Circuit Switching

One of the defining characteristics of the traditional telephone network was circuit switching.

Suppose Alice called Bob.

The telephone network established a communication path through the switching system:

Alice’s Telephone

↓

Local Telephone Exchange

↓

Telephone Trunks

↓

Remote Telephone Exchange

↓

Bob’s Telephone

Network resources were allocated to that conversation while the call was in progress.

That is fundamentally different from the way ordinary Internet packets are handled.

In a packet-switched network, packets belonging to many users normally share network links.

In a traditional circuit-switched telephone system, resources were associated with an established call.

---

A Dedicated Circuit Did Not Mean a Dedicated Physical Wire

There is an important misconception here.

A circuit-switched telephone call did not necessarily mean that one individual copper wire ran all the way from one caller to another.

The network could multiplex many calls over shared transmission facilities.

What was dedicated or reserved was a logical communication channel or network capacity associated with the call.

So:

Dedicated circuit

does not necessarily mean:

Dedicated physical cable from caller A to caller B

The physical infrastructure could be shared while the switching system maintained a logical circuit for each conversation.

---

What Is a Data Network?

A useful modern definition is:

A data network is a network designed to exchange digital data between devices using communication protocols. The data may represent text, numbers, files, images, audio, voice, video, sensor measurements, or other information.

Examples of devices participating in data networks include:

• desktop computers;
• laptops;
• smartphones;
• servers;
• routers;
• sensors;
• industrial controllers;
• cameras;
• IoT devices.

Examples of data-network activities include:

• loading a webpage;
• sending email;
• transferring files;
• querying a database;
• streaming video;
• sending sensor measurements;
• making a VoIP call.

That last example illustrates why the old distinction between voice and data has become increasingly difficult to maintain.

---

Isn’t Voice Also Data?

Today, yes.

In the broad technical meaning of the term, digitally represented voice is data.

Consider a modern telephone call.

Human speech can go through:

Voice

↓

Microphone

↓

Analog electrical representation

↓

Analog-to-digital conversion

↓

Digital audio samples

↓

Voice codec

↓

Compressed digital data

↓

Packets

↓

Communication network

Once voice has been sampled, digitally encoded, and packetized, the network is moving digital data representing speech.

The same principle applies to images and video.

Image

→ pixel data → encoded file → bits

Video

→ image/audio samples → compression → digital stream → packets

Text

→ character encoding → numbers → bits

Therefore:

Voice, images, video, and text can all become data when represented digitally.

---

Why, Then, Do We Still Say “Voice and Data”?

This terminology comes largely from telecommunications history.

Traditionally:

Voice traffic

meant telephone conversations.

Data traffic

usually meant non-voice information exchanged by computers.

So engineers commonly distinguished:

Voice network

from

Data network

because these could genuinely be different networks using different technologies.

Today the expression “voice and data” is still widely used, but technically voice can itself be digital data.

The word data is therefore being used in two slightly different ways.

Broad technical meaning

Data includes digitally represented:

• voice;
• text;
• images;
• video;
• files;
• measurements.

Traditional telecommunications meaning

“Data” often meant:

computer/non-voice traffic

This historical usage explains much of the terminology that remains today.

---

Telephone Networks Did Not Jump Directly From Analog to the Internet

The evolution of telephony occurred in several important stages.

A simplified progression is:

Analog circuit switching

↓

Digital circuit switching

↓

Hybrid telephone/IP networks

↓

Packet-switched IP voice networks

Understanding the middle stage is particularly important.

---

Stage 1: Analog Telephone Network

Early telephone systems could be simplified as:

Voice

↓

Analog telephone

↓

Analog local line

↓

Circuit-switched telephone network

↓

Analog local line

↓

Telephone

↓

Voice

Much of the communication path was analog.

---

Stage 2: The Telephone Network Became Digital

Telephone companies gradually digitized their internal networks.

The customer might still have an ordinary analog telephone:

Analog Telephone

↓

Copper Telephone Pair

but the telephone exchange could convert the voice into digital form.

The path then became approximately:

Analog Telephone

↓

Analog Local Loop

↓

Telephone Exchange

↓

Analog-to-Digital Conversion

↓

Digital Telephone Network

↓

Digital-to-Analog Conversion

↓

Analog Local Loop

↓

Analog Telephone

This was a major technological change.

But the network could still remain circuit switched.

---

Digital Does Not Automatically Mean Internet

This distinction is critical.

A network can be:

digital + circuit switched

or:

digital + packet switched

Traditional digital telephone systems commonly used technologies such as PCM and Time Division Multiplexing (TDM).

Multiple telephone conversations could be digitized and carried together across high-capacity digital transmission systems.

Yet each call still behaved as a circuit-switched telephone connection.

Therefore:

Telephone networks became digital long before they became packet-switched IP networks.

---

Stage 3: IP Entered the Middle of the Telephone Network

The next major change was the use of packet networks to carry telephone calls.

A customer could still use a traditional telephone interface while the carrier converted the voice to IP internally.

For example:

Traditional Telephone

↓

Telephone Network

↓

Media Gateway

↓

IP Packets

↓

Carrier IP Network

↓

Media Gateway

↓

Telephone Network

↓

Traditional Telephone

The caller might never know that part of the conversation travelled as IP packets.

The U.S. Federal Communications Commission has used the expression “IP-in-the-middle” for arrangements in which traditional telephone service uses IP technology within portions of the network.

This demonstrates an important principle:

A traditional-looking telephone service can use modern packet networking internally without changing the user’s telephone experience.

---

IP Network Does Not Necessarily Mean the Public Internet

Another common misconception is:

If telephone calls use IP, they must travel over the public Internet.

That is not necessarily true.

A telecommunications company can operate its own managed IP network.

The carrier might use:

• private IP backbones;
• MPLS networks;
• fiber networks;
• managed Ethernet networks;
• IMS infrastructure;
• private data centres.

Therefore:

IP network ≠ public Internet

The Internet uses IP, but IP technology can also be used inside completely private networks.

A telephone call may consequently travel as IP packets without being routed over the ordinary public Internet.

---

Stage 4: Modern Packet-Based Voice

Modern telephone systems increasingly treat voice as packetized digital information.

A simplified modern communication path is:

Human Voice

↓

Microphone

↓

Digital Samples

↓

Voice Codec

↓

Packets

↓

IP Network

↓

Packets

↓

Voice Decoder

↓

Speaker

↓

Human Voice

The network does not need to transport one continuous analog waveform from one telephone all the way to another.

Instead, it transports digital information representing the voice.

---

Modern Landline Phones Can Still Look Completely Traditional

An interesting aspect of modern telecommunications is that the telephone itself does not necessarily have to be modern.

Imagine an ordinary analog telephone plugged into a residential gateway.

The architecture can be:

Analog Telephone

↓

Telephone Port on Gateway/ONT

↓

Analog-to-Digital Conversion

↓

Voice Encoding

↓

IP Packets

↓

Fiber/Cable/Broadband Network

↓

Carrier Voice System

The gateway handles the technological conversion.

The telephone itself may know nothing about:

• IP;
• Ethernet;
• fiber;
• packet switching;
• Internet protocols.

To the person using it, it is simply a telephone.

---

The Same Fiber Can Carry Internet, Television, and Telephone Services

Modern broadband infrastructure illustrates network convergence.

A single fiber connection entering a home may support:

Internet

Telephone

Television

The services can still be logically separated and managed differently, but they can share substantial portions of the same physical infrastructure.

This contrasts sharply with older networks in which separate infrastructures were often constructed for separate services.

---

Cellular Voice Went Through a Similar Transformation

Mobile telephone networks experienced their own transition.

Earlier cellular systems were strongly oriented toward circuit-switched telephone calls.

Later mobile generations supported both:

Circuit-Switched Domain → Voice

and

Packet-Switched Domain → Data

A simplified 3G architecture could therefore have separate mechanisms for ordinary voice and Internet data.

LTE changed the model significantly.

LTE was designed as a packet-oriented system.

Voice increasingly moved to technologies such as:

VoLTE — Voice over LTE

and, with 5G:

VoNR — Voice over New Radio

Now the voice itself can be handled as an application/service on top of a sophisticated packet-based telecommunications infrastructure.

---

A Modern Mobile Call

A modern mobile voice call can conceptually look like:

Human Voice

↓

Smartphone

↓

Voice Codec

↓

Digital Packets

↓

4G LTE or 5G Radio

↓

Base Station

↓

Mobile Packet Core

↓

IMS / Voice Platform

↓

Remote Network

↓

Recipient’s Phone

This is quite different from the traditional analog telephone model.

---

The Telephone Network Has Become Increasingly Software-Based

The old telephone network depended heavily on specialized switching equipment.

Modern telephone systems increasingly depend on software.

Examples include:

• SIP signaling;
• IMS;
• software-controlled call routing;
• subscriber databases;
• authentication systems;
• virtual network functions;
• cloud infrastructure;
• software-defined networking;
• IP routers;
• application servers.

The physical network remains essential.

It still requires:

• fiber;
• radio towers;
• antennas;
• switches;
• routers;
• data centres;
• submarine cables;
• satellites.

But much of the intelligence that decides how calls are established, routed, authenticated, billed, and terminated is now implemented in software.

---

Voice Networks and Data Networks Have Converged

Historically, we might have drawn:

Telephone → Voice Network → Telephone

and separately:

Computer → Data Network → Computer

Today, a more realistic picture is:

Voice

Video

Images
→ Digital Data → Packet Network
/
Text
/
Applications

The same network infrastructure can support many types of information.

This process is generally called network convergence.

---

Does a Dedicated Telephone Network Still Exist Today?

Yes—but increasingly as legacy or specialized infrastructure, rather than as the dominant architecture.

It is important to distinguish between:

1. a network built primarily for telephone service;
2. a copper telephone access line;
3. circuit-switched PSTN infrastructure;
4. a network that literally carries only voice.

The fourth category is now relatively uncommon because even traditional telephone infrastructure has frequently been adapted to carry additional services.

Nevertheless, several concrete examples show that legacy telephone-specific infrastructure remains operational in 2026.

---

United Kingdom: BT’s PSTN Still Exists Until January 2027

As of September 2026, the United Kingdom is an unusually clear example of a country in the final stage of PSTN retirement.

BT plans to retire its traditional Public Switched Telephone Network on:

January 31, 2027.

UK regulator Ofcom stated in January 2026 that BT’s PSTN remains in service while customers are migrated from analog landlines to digital/VoIP services. Ofcom describes the technology as beyond its intended lifespan and increasingly unreliable.

BT reported in March 2026 that approximately 80% of its voice customers had already moved to All-IP services as of January 2026, but thousands of organizations continued to depend on PSTN-based services.

So in September 2026:

Traditional PSTN infrastructure still exists in the UK, but it is scheduled for retirement on January 31, 2027.

The migration is:

Traditional PSTN

↓

Digital landline

↓

VoIP / All-IP infrastructure

Ofcom also notes that other UK operators, including Virgin Media O2 and KCOM, operate legacy networks and intend to retire them, although their final shutdown dates are not necessarily identical to BT’s.

---

Australia: Telstra Still Supports Copper PSTN Voice

Australia provides another particularly interesting example.

Telstra currently states that although it is retiring certain older access services, it is not exiting copper PSTN voice services in the affected context. Customers who want to retain copper PSTN telephone service can purchase a standalone Home or Business Phone service.

Telstra also operates a specialized Customer Access Network Radio (CAN Radio) voice service for people in remote areas.

These systems use radio links connecting customers in isolated locations to telephone infrastructure.

Telstra plans to exit CAN Radio by:

November 16, 2027.

This is an excellent real-world example of a specialized network that remains primarily associated with traditional voice service.

So, even in a country with highly developed modern broadband infrastructure, legacy voice-specific systems can continue because of geography, existing equipment, and the needs of remote customers.

---

France: Orange Is Retiring Its Historic Copper Network Through 2030

France illustrates a slower nationwide transition.

France’s telecom regulator, Arcep, describes Orange’s copper infrastructure as the historic network used for fixed telephone service and later broadband technologies such as ADSL, SDSL, and VDSL.

Orange is progressively shutting down this network, with full technical closure planned by the end of 2030.

The retirement is happening area by area.

Technical shutdown began in some locations in 2025, while successive groups of communities will migrate until the nationwide shutdown is completed.

France also demonstrates why copper network and voice-only network should not be treated as synonyms.

The same copper infrastructure that historically carried telephone service was later also used for DSL Internet access.

So it is legacy telecommunications infrastructure, but not necessarily a network carrying only voice.

---

United States: Legacy Copper Telephone Infrastructure Is Still Being Retired

The United States is also still undergoing its transition away from legacy telephone infrastructure.

In March 2026, the Federal Communications Commission adopted measures intended to accelerate migration from aging copper telephone networks toward modern IP-based services.

The FCC explicitly described the transition as moving away from:

legacy copper telephone networks

toward:

IP-based services.

The fact that this regulatory transition remains active in 2026 demonstrates that traditional copper-based telephone infrastructure has not vanished across the United States.

Migration happens unevenly because the country contains:

• dense urban networks;
• suburban systems;
• rural telephone companies;
• remote communities;
• specialized business services;
• legacy alarm and emergency systems.

---

Canada: Legacy Fixed Infrastructure and 3G Are Also Being Retired

Canada is undergoing the same general technological transition.

A 2026 CRTC decision concerning Bell Canada notes that legacy copper infrastructure is being replaced by newer fiber-based infrastructure.

Bell Canada is also withdrawing or destandardizing several older services as the equipment becomes difficult to maintain. For example, a July 2026 CRTC decision describes Bell’s Megalink service as increasingly obsolete and preserves it for existing customers while supporting migration to alternatives.

The transition is equally visible in Canadian mobile networks.

Bell continues to operate its national 3G/HSPA network in most of Canada as of September 2026, but plans to discontinue it nationally on:

March 1, 2027.

Manitoba’s Bell 3G/HSPA service was already discontinued on December 31, 2025.

Bell specifically warns that devices dependent on 3G may lose:

• voice;
• text;
• data;
• 9-1-1 access.

Devices must move toward LTE/VoLTE or newer technologies.

This is another important distinction: 3G is not a dedicated voice-only network. It carries voice, text, and data. But older mobile voice architectures are nevertheless being replaced by packet-based LTE/5G and VoLTE systems.

---

The UK Has Already Shut Down 3G Mobile Networks

The transition is occurring at different speeds in different countries.

As of May 2026, Ofcom reports that all four major UK mobile operators—

• Vodafone;
• EE;
• Three;
• O2—

have completed their 3G shutdowns.

The UK is now progressively retiring 2G as well.

This shows that there is no single worldwide date when traditional voice technologies disappear.

Different countries and operators are at different points in the migration.

---

The Meaning of “Voice Network” Is Changing

Historically, a voice network meant something close to:

A dedicated circuit-switched telecommunications network designed primarily for telephone voice communication.

Today the term may instead describe:

The systems, applications, signaling platforms, and network resources used to deliver voice services, even when the underlying transport is a shared packet-switched IP network.

This distinction is important.

A modern organization can speak of its “voice network” even if that network consists of:

• IP phones;
• Ethernet switches;
• SIP servers;
• routers;
• fiber;
• cloud PBXs;
• Internet connections.

The service is still voice, but the underlying infrastructure is a data network.

---

A Better Modern Definition of Voice Network

A modern definition would be:

Voice network — a communication network or system designed primarily to provide voice communication services. Traditional voice networks were generally circuit-switched, while modern voice services are increasingly digitized, packetized, and transported over shared IP-based networks.

This preserves both the historical and current meanings.

---

A Better Modern Definition of Data Network

A corresponding definition is:

Data network — a network designed to exchange digital data between devices using communication protocols. The data may represent text, files, images, audio/voice, video, sensor information, or other digital content.

And the relationship between them has changed.

Historically:

Voice network ≠ Data network

Today:

Voice service → frequently carried over a data network

---

How Voice Works on a Modern Data Network

Suppose someone makes a modern VoIP or VoLTE telephone call.

The process can be simplified as:

1. Speech is produced

The person creates an analog sound wave.

2. The microphone captures it

The microphone converts sound pressure into an electrical signal.

3. The signal is digitized

An analog-to-digital converter samples the waveform and creates numerical values.

4. A codec compresses the voice

A voice codec encodes the speech efficiently.

5. Voice becomes packet data

The encoded information is placed into packets.

6. Packets travel through the network

Those packets may travel over:

• Ethernet;
• Wi-Fi;
• LTE;
• 5G;
• fiber;
• cable;
• carrier IP networks;
• private networks;
• portions of the Internet.

7. The receiver reconstructs the voice

The destination device:

receives packets

↓

extracts encoded audio

↓

decodes the audio

↓

creates a speaker signal

↓

produces sound

The recipient hears the conversation.

---

The Physical Signal Can Change Several Times

The voice information may also change physical representation during its journey.

For example:

Human voice

↓

electrical signal inside smartphone

↓

digital data

↓

5G radio wave

↓

electrical representation at base station

↓

optical signal through fiber

↓

electrical signal in router

↓

optical signal

↓

radio signal

↓

receiving smartphone

↓

sound

The information remains the conversation.

The data representation allows computers and network devices to process it.

The physical signal changes according to the medium.

---

The Network No Longer Needs to Know That the Payload Is “A Telephone Call”

This is one of the profound consequences of packet networking.

At many parts of a modern IP network, routers are primarily concerned with things such as:

• source and destination IP addresses;
• routing tables;
• packet headers;
• traffic policies;
• quality-of-service markings.

A router does not need to understand the meaning of every spoken word.

To much of the network, voice is simply:

packets that must be delivered.

This allows a shared infrastructure to carry:

Voice

Video

Web

Email

Files

Cloud Applications

IoT Data

simultaneously.

---

Why Keep Separate Voice Services at All?

If voice is now data, one might ask why telecommunications companies still treat telephone service separately.

There are several reasons.

Telephone services have special requirements involving:

Reliability

People expect telephone calls to work consistently.

Quality of Service

Real-time voice is sensitive to:

• delay;
• jitter;
• packet loss.

Emergency Calling

Telephone networks must support services such as:

9-1-1 / 112 / 999

depending on the country.

Telephone Numbers

The global telephone numbering system must still work.

Interconnection

Calls need to move between different carriers and different technologies.

Regulatory Requirements

Telephone providers may have obligations that ordinary Internet applications do not.

Power and Resilience

Traditional copper telephone systems sometimes supplied power from the telephone exchange.

Digital phones connected through home broadband equipment may depend on electricity at the customer’s premises, which creates additional requirements for backup power and vulnerable customers.

This issue is one reason regulators such as Ofcom are closely supervising the migration from traditional PSTN lines to digital telephony.

---

Voice Has Become an Application on the Network

Perhaps the simplest way to understand the modern transformation is this:

Historically:

The network was built specifically for telephone voice.

Today:

Voice is increasingly one service or application running over a general-purpose digital packet network.

Compare:

Traditional model

Telephone

↓

Voice Network

↓

Telephone

Modern model

Telephone/Smartphone

↓

Voice Application/Service

↓

IP Packet Network

↓

Voice Application/Service

↓

Telephone/Smartphone

That is a fundamental architectural change.

---

From Separate Networks to Converged Networks

Historically, an organization might have needed different infrastructure for:

Telephone

Computer networking

Television/video

Today, the same Ethernet and IP infrastructure can potentially support all three.

For example:

Ethernet/IP Network

├── Computers

├── IP Phones

├── Wi-Fi

├── Video conferencing

├── Security cameras

├── Servers

└── Internet access

This is generally called a converged network.

---

Voice Networks Are Disappearing as Separate Transport Networks, Not as Voice Services

This distinction is worth emphasizing.

Voice communication is certainly not disappearing.

People still make billions of telephone and Internet voice calls.

What is disappearing is the need for a completely separate physical network whose primary job is transporting telephone voice.

Instead:

Voice service remains

while:

Dedicated legacy transport infrastructure gradually disappears.

Voice increasingly uses the same:

• fiber;
• Ethernet;
• IP;
• radio;
• routing;
• data-centre infrastructure

used by other digital services.

---

Where Things Stand in September 2026

The worldwide transition can be summarized approximately as follows:

Technology	Status in 2026
Analog circuit-switched PSTN	Still exists, but rapidly declining
Copper telephone access	Still widely present in some countries, being retired
Digital TDM telephone networks	Legacy/declining
2G circuit-switched mobile voice	Still exists in some countries; retirement underway
3G voice/data	Being retired rapidly
VoIP	Mainstream
VoLTE	Mainstream mobile voice technology
5G/VoNR	Growing
Fiber-based voice	Increasing
Cloud/IP PBX	Mainstream business technology
Shared IP transport	Dominant direction

The timelines differ significantly by operator and country.

For example:

UK / BT: PSTN retirement planned for January 31, 2027.

France / Orange: historic copper network planned for full technical retirement by the end of 2030.

Australia / Telstra: some standalone copper PSTN voice remains available; CAN Radio voice service retirement is planned for November 16, 2027.

Canada / Bell: nationwide 3G/HSPA shutdown planned for March 1, 2027.

United States: the FCC was still actively changing rules in March 2026 to accelerate migration away from legacy copper telephone networks to IP-based services.

The direction is clear even though the transition is not complete everywhere:

Dedicated legacy telephone infrastructure

↓

Digital converged infrastructure

↓

All-IP telecommunications

---

Final Definitions

After considering both historical and modern systems, these definitions are useful.

Voice Network

A voice network is a communication network or system designed primarily to provide voice communication services. Traditional voice networks were typically circuit-switched and originally carried analog telephone voice signals; modern voice services are increasingly digital, packet-switched, and transported over shared IP-based infrastructure.

Data Network

A data network is a network designed to exchange digital data between devices using communication protocols. The data may represent text, numbers, files, images, audio/voice, video, sensor measurements, or other digital information.

Converged Network

A converged network is a common digital network infrastructure capable of carrying multiple types of services, such as voice, video, and computer/application data.

---

Final Thought

The phrase “voice network” once described a fundamentally different infrastructure from a computer data network.

That was a world in which:

Telephone → telephone network

and:

Computer → computer network

were largely separate.

That world has changed.

Telephone voice first moved from analog transmission toward digital circuit switching.

Then telephone carriers adopted packet networks and IP transport.

Mobile networks progressed from circuit-switched cellular voice toward VoLTE and 5G packet-based communication.

Fiber, broadband, carrier IP networks, data centres, and software platforms increasingly replaced infrastructure dedicated solely to carrying telephone calls.

Yet the transition is still happening.

Legacy PSTN, copper telephone, TDM, 2G/3G, and specialized voice systems continue to operate in some places in 2026, including parts of the United Kingdom, France, Australia, Canada, and the United States.

The most accurate way to describe the transformation is therefore not:

“Voice networks no longer exist.”

Instead:

Traditional dedicated voice networks are gradually being retired, while voice itself is becoming a digital service carried over converged packet-switched IP networks.

The conversation is still voice.

What changed is the network underneath it.

References and Further Reading

1. Ofcom — PSTN switch-off and migration to Voice over IP (VoIP), January 29, 2026. Ofcom states that BT plans to retire its PSTN by January 31, 2027 and discusses the migration of analog telephone customers to digital services.
2. BT — Digital Switchover / All-IP. BT describes the replacement of its traditional analog telephone network by digital Voice over IP services.
3. BT Business — Final phase of the PSTN switchover, March 20, 2026. BT reported that around 80% of its voice customers had moved to All-IP services as of January 2026 while thousands of organizations remained on PSTN-based services.
4. Ofcom — The future of landline calls. Information about BT, Openreach, Virgin Media O2, KCOM, digital landlines, and the transition away from legacy networks.
5. Australian Telstra — CAN Radio and ADSL Service Exit. Telstra describes remote CAN Radio voice systems, their planned November 16, 2027 retirement, and continued availability of standalone copper PSTN voice services in relevant areas.
6. Arcep — Closure of the French copper network. France’s regulator describes Orange’s historic copper telecommunications network and the transition toward fiber.
7. Orange Wholesale France — Copper Network. Orange describes the progressive closure of its nationwide copper network, with technical shutdown expected to be completed by the end of 2030.
8. Federal Communications Commission — Accelerating Network and Service Modernization, March 26, 2026. The FCC discusses the continuing U.S. transition from aging legacy copper telephone networks to IP-based services.
9. Bell Canada — 3G/HSPA Network Discontinuation. Bell states that its 3G/HSPA network is scheduled to cease nationally on March 1, 2027, following the December 31, 2025 shutdown in Manitoba.
10. CRTC — Telecom Order 2026-28. The CRTC discusses Bell Canada’s declining legacy landline services and the replacement of legacy copper infrastructure with fiber-based infrastructure.
11. Ofcom — 2G and 3G switch-off. Ofcom confirms that all four major UK mobile operators have completed their 3G shutdowns and that 2G retirement is progressing.

Understanding the Electromagnetic Spectrum, Radio Frequencies, and Wi-Fi Ref: AI Tools as is

When we talk about Wi-Fi, cellular networks, radio, Bluetooth, satellite communication, infrared, visible light, X-rays, or gamma rays, we are talking about different parts of the same broad phenomenon: the electromagnetic spectrum.

The main difference between these forms of electromagnetic radiation is their frequency and wavelength.

Understanding the spectrum makes it much easier to understand how wireless communication works.

What Is the Electromagnetic Spectrum?

The electromagnetic spectrum is the full range of electromagnetic waves.

These waves can travel through space and carry energy.

They include:

• radio waves;
• microwaves;
• infrared;
• visible light;
• ultraviolet;
• X-rays;
• gamma rays.

A simplified spectrum looks like this:

Low frequency → High frequency

Radio → Microwave → Infrared → Visible Light → Ultraviolet → X-ray → Gamma Ray

As frequency increases, wavelength decreases.

That relationship is expressed by:$$c = f\lambda$$

where:

• $c$ = speed of light;
• $f$ = frequency;
• $\lambda$ = wavelength.

Because the speed of light is approximately constant, frequency and wavelength move in opposite directions.

Higher frequency means shorter wavelength.

Lower frequency means longer wavelength.

---

What Does Frequency Mean?

Frequency tells us how many wave cycles occur every second.

The unit is the hertz, abbreviated Hz.

For example:

• 1 Hz = 1 cycle per second
• 1 kHz = 1,000 cycles per second
• 1 MHz = 1,000,000 cycles per second
• 1 GHz = 1,000,000,000 cycles per second
• 1 THz = 1,000,000,000,000 cycles per second

Therefore, a Wi-Fi signal at approximately 2.4 GHz oscillates about:

2.4 billion times per second.

A 5 GHz signal oscillates approximately:

5 billion times per second.

This does not mean that 5 GHz Wi-Fi necessarily transfers exactly twice as much data as 2.4 GHz Wi-Fi. Frequency is only one part of the communication system.

Bandwidth, modulation, channel width, signal quality, interference, antenna design, coding, and protocol design also matter.

---

The Overall Electromagnetic Spectrum

The electromagnetic spectrum is much larger than the radio spectrum.

Approximate ranges are:

Region	Approximate Frequency
Radio	about 3 kHz – 300 GHz
Infrared	about 300 GHz – 400 THz
Visible light	about 400 – 790 THz
Ultraviolet	about 790 THz and above
X-rays	much higher frequencies
Gamma rays	extremely high frequencies

The exact boundaries are not perfectly fixed. Different scientific and engineering references may define them slightly differently.

The important point is that radio is only one portion of the complete electromagnetic spectrum.

---

What Is the Radio Spectrum?

For communications engineering, the conventional radio-frequency spectrum is approximately:

3 kHz to 300 GHz

This is an enormous range.

Radio frequencies are divided into bands.

Band	Frequency Range
VLF – Very Low Frequency	3–30 kHz
LF – Low Frequency	30–300 kHz
MF – Medium Frequency	300 kHz–3 MHz
HF – High Frequency	3–30 MHz
VHF – Very High Frequency	30–300 MHz
UHF – Ultra High Frequency	300 MHz–3 GHz
SHF – Super High Frequency	3–30 GHz
EHF – Extremely High Frequency	30–300 GHz

Different technologies use different portions of this spectrum.

---

Examples of Technologies Across the Radio Spectrum

Different radio frequencies are suitable for different purposes.

Examples include:

• AM radio
• FM radio
• television broadcasting
• maritime communication
• aviation communication
• cellular networks
• GPS
• Wi-Fi
• Bluetooth
• satellite communication
• radar

For example:

AM radio uses relatively low radio frequencies.

FM radio uses higher frequencies.

Cellular networks use many frequencies ranging from hundreds of MHz into several GHz.

Wi-Fi commonly operates around 2.4 GHz, 5 GHz, and 6 GHz.

---

Where Does Wi-Fi Fit?

Wi-Fi uses radio waves.

Modern Wi-Fi primarily operates in three frequency regions:

• 2.4 GHz
• 5 GHz
• 6 GHz

These are all within the radio-frequency spectrum.

A simplified view is:

Radio spectrum: 3 kHz ——————————————————— 300 GHz

Within that enormous range:

Wi-Fi occupies relatively small bands around 2.4, 5, and 6 GHz.

So Wi-Fi does not have access to the entire radio spectrum.

It uses only specific ranges allocated for wireless networking.

---

2.4 GHz Wi-Fi

The 2.4 GHz band is one of the most familiar Wi-Fi frequency ranges.

A typical Wi-Fi allocation is roughly around:

2.4 GHz to 2.4835 GHz

This band is also used by many other devices.

Examples include:

• Bluetooth;
• wireless keyboards;
• wireless mice;
• some smart-home devices;
• microwave ovens;
• some cordless equipment.

Because so many technologies share this part of the spectrum, congestion and interference can occur.

Advantages of 2.4 GHz

Lower frequencies generally have longer wavelengths.

This often gives 2.4 GHz Wi-Fi:

• better wall penetration;
• somewhat longer practical range;
• better coverage through obstacles.

Disadvantages

It also tends to have:

• fewer non-overlapping Wi-Fi channels;
• more interference;
• more congestion.

---

5 GHz Wi-Fi

Wi-Fi can also operate in several portions of the 5 GHz range.

Compared with 2.4 GHz, 5 GHz generally provides:

• more available channels;
• wider channels;
• less congestion in many environments;
• potentially higher data rates.

However, higher frequencies generally experience more attenuation through walls and obstacles.

Therefore:

5 GHz may provide better speed, but often shorter practical range than 2.4 GHz.

This is not an absolute rule, because actual performance depends heavily on:

• antenna placement;
• transmit power;
• building materials;
• interference;
• access-point design;
• device capabilities.

---

6 GHz Wi-Fi

Newer Wi-Fi technologies can also use spectrum around 6 GHz.

This is associated particularly with:

• Wi-Fi 6E;
• Wi-Fi 7.

The 6 GHz band provides substantially more spectrum for wireless LAN operation in regions where regulators permit it.

More spectrum means the possibility of:

• more channels;
• wider channels;
• less congestion;
• higher throughput.

However, like 5 GHz, 6 GHz signals generally do not penetrate obstacles as well as lower-frequency 2.4 GHz signals.

---

Why Does Wi-Fi Need Different Frequency Bands?

There is a basic engineering trade-off.

Lower frequencies often provide:

better propagation and greater coverage

while higher frequencies can provide access to:

more bandwidth and potentially greater capacity.

A simplified comparison is:

Wi-Fi Band	Range	Capacity	Wall Penetration
2.4 GHz	Generally better	Lower	Generally better
5 GHz	Moderate	Higher	Moderate
6 GHz	Generally shorter	Very high potential	Generally lower

These are general tendencies, not fixed guarantees.

---

Frequency Is Not the Same as Bandwidth

This is one of the most important distinctions.

Frequency tells us where a signal is located in the spectrum.

Bandwidth tells us how much spectrum a signal occupies.

For example, suppose a Wi-Fi channel is centred somewhere near 5 GHz.

That does not mean the channel has 5 GHz of bandwidth.

It may use a channel width such as:

• 20 MHz;
• 40 MHz;
• 80 MHz;
• 160 MHz;
• or even wider channels in newer standards.

So:

5 GHz = approximate operating frequency

while:

80 MHz = possible channel bandwidth.

These are very different concepts.

---

An Analogy: Radio Spectrum as a Highway

The spectrum can be imagined as a very large highway.

The frequency tells you where on the highway you are.

The bandwidth tells you how many lanes you are using.

For example:

A wireless system might operate near:

5 GHz

and use:

80 MHz of bandwidth.

The 5 GHz number tells us the location.

The 80 MHz number tells us how wide a portion of spectrum the communication occupies.

A wider channel can potentially carry more information, just as a wider highway can potentially carry more traffic.

---

Why Can’t Wi-Fi Use Any Frequency It Wants?

Because the radio spectrum is a shared resource.

If every device transmitted at any frequency and power level it wanted, wireless communications would interfere with one another.

Governments and international organizations therefore regulate the use of spectrum.

Different bands may be allocated for:

• broadcasting;
• cellular communication;
• satellite communication;
• aviation;
• public safety;
• navigation;
• military systems;
• amateur radio;
• Wi-Fi and other unlicensed technologies.

Some spectrum is licensed.

For example, mobile network operators commonly pay for rights to use particular frequency bands.

Other spectrum is made available for unlicensed use, subject to technical rules.

Wi-Fi primarily operates in unlicensed spectrum.

“Unlicensed” does not mean unregulated.

It means users generally do not need an individual radio licence to operate compliant equipment.

---

Wi-Fi and Cellular Networks Both Use Radio Waves

It is easy to think of Wi-Fi and cellular networks as fundamentally different physical technologies.

At the radio level, however, both use electromagnetic radio waves.

The difference is largely in:

• frequencies;
• channel organization;
• modulation;
• power;
• network architecture;
• access control;
• spectrum licensing;
• mobility management.

A Wi-Fi connection may look like:

Phone → Wi-Fi Access Point → Router → Internet

A cellular connection may look like:

Phone → Cellular Base Station → Mobile Core Network → Internet

Both begin with radio communication.

---

What About Bluetooth?

Bluetooth also uses radio waves.

Most conventional Bluetooth communication operates in the approximately 2.4 GHz ISM band.

That means Wi-Fi and Bluetooth may sometimes operate within overlapping portions of the spectrum.

Modern systems use various techniques to reduce interference.

This is another example of why spectrum management matters.

---

What Is a Wavelength?

Frequency can also be understood through wavelength.

The approximate wavelength is:$$\lambda = \frac{c}{f}$$

where $c$ is approximately:$$3 \times 10^8 \text{ metres per second}$$

For 2.4 GHz:$$\lambda \approx \frac{3\times10^8}{2.4\times10^9}$$

which is about:

12.5 cm

For 5 GHz:$$\lambda \approx 6 \text{ cm}$$

For 6 GHz:$$\lambda \approx 5 \text{ cm}$$

So higher-frequency radio signals have shorter wavelengths.

This affects:

• antenna size;
• propagation;
• diffraction;
• penetration;
• reflection;
• radio-system design.

---

Why Do Lower Frequencies Often Travel Farther?

Lower-frequency signals generally have longer wavelengths.

Longer wavelengths can often:

• diffract around obstacles more effectively;
• penetrate some materials better;
• experience different propagation losses.

This is one reason lower-frequency cellular spectrum can be valuable for wide-area coverage.

Higher-frequency spectrum can offer much more bandwidth, but often requires denser infrastructure.

This trade-off is seen in both cellular and Wi-Fi networks.

---

Is Microwave Different From Radio?

The terminology can sometimes be confusing.

Microwaves are generally considered a subset of radio waves.

A commonly used microwave range is approximately:

300 MHz to 300 GHz

Therefore:

• 2.4 GHz Wi-Fi is microwave radio;
• 5 GHz Wi-Fi is microwave radio;
• 6 GHz Wi-Fi is microwave radio.

Calling something a microwave signal does not mean it is fundamentally different from radio.

It simply refers to a particular part of the radio spectrum.

---

What Is Millimeter Wave?

Millimeter wave generally refers to very high radio frequencies with wavelengths measured in millimetres.

A commonly used range is approximately:

30 GHz to 300 GHz

These frequencies can support very large bandwidths.

They are used or investigated for applications such as:

• high-capacity wireless communication;
• radar;
• satellite systems;
• some 5G deployments.

Their disadvantage is that propagation becomes more difficult.

Buildings, walls, foliage, and even atmospheric effects can have greater impact.

---

The Big Picture

The easiest way to understand wireless communication is to see Wi-Fi as one tiny part of a much larger spectrum.

The full electromagnetic spectrum includes:

Radio → Infrared → Visible Light → Ultraviolet → X-rays → Gamma Rays

The radio portion is approximately:

3 kHz to 300 GHz

Within that radio spectrum are many services:

Broadcasting → Cellular → GPS → Wi-Fi → Bluetooth → Satellite → Radar

Wi-Fi itself mainly uses:

2.4 GHz + 5 GHz + 6 GHz

These numbers describe the frequency of the electromagnetic waves carrying the information.

They do not directly tell us the Internet speed.

The actual data rate depends on many additional factors, including bandwidth, modulation, signal quality, channel conditions, number of antennas, coding, and the Wi-Fi standard being used.

Final Thought

When a phone or laptop sends information over Wi-Fi, there is no invisible cable connecting it to the access point.

Instead, electronic circuits convert digital information into a carefully controlled radio-frequency signal.

That signal travels through the surrounding electromagnetic field as radio waves.

The receiving device detects those waves, extracts the encoded information, and reconstructs the digital data.

So when we say:

“My laptop is connected through Wi-Fi,”

what is physically happening is:

digital information → radio-frequency electromagnetic signal → radio waves through space → receiver → digital information.

Wi-Fi is therefore not separate from radio technology.

Wi-Fi is one specialized way of using a small portion of the radio spectrum to create a computer network.