From landline telephony to 5G: Understanding the evolution and architecture of telecommunications networks
The Public Switched Telephone Network (PSTN) laid the foundation for global telecommunications, while mobile networks revolutionized how we communicate. This topic explores the evolution from traditional circuit-switched telephony to modern 5G packet-switched networks.
Understanding the transition from PSTN to mobile networks is crucial for telecommunications professionals. The PSTN, developed over more than a century, represents one of humanity's most complex engineering achievements—a global network connecting billions of endpoints through hierarchical switching exchanges. While PSTN relied on dedicated circuit-switched connections that reserved bandwidth for the entire duration of a call, modern mobile networks use packet-switched technology that efficiently shares network resources among millions of users simultaneously.
Mobile networks have undergone five major generational shifts, each bringing transformative improvements. From the analog voice-only 1G systems of the 1980s to today's 5G networks supporting ultra-low latency applications like autonomous vehicles and remote surgery, each generation represented a 10-100x improvement in capabilities. This evolution wasn't merely about faster speeds—it fundamentally changed how we use telecommunications, enabling smartphones, mobile internet, video streaming, IoT devices, and emerging technologies like augmented reality.
This comprehensive guide covers the architectural principles, technological innovations, and operational concepts that define both traditional and modern telecommunications networks. You'll gain deep insights into how calls are routed through PSTN hierarchies, how cellular networks manage millions of mobile devices, how spectrum allocation enables wireless communications, and how technologies like handoff and roaming provide seamless connectivity as users move across coverage areas.
PSTN is the traditional circuit-switched telephone network that has served as the backbone of voice communications for over a century. It establishes dedicated communication paths between callers for the duration of a call.
The fundamental principle of PSTN is circuit switching, where a dedicated physical or virtual circuit is established between two endpoints for the entire duration of a call. When you dial a phone number, switches along the path from source to destination create a continuous connection by physically or logically connecting wires or time slots. This connection remains exclusively reserved for your call until one party hangs up, ensuring consistent quality and zero packet loss—critical for real-time voice communication.
Unlike modern packet-switched networks where data is broken into packets that can take different paths, circuit-switched calls follow a predetermined route. This architecture guarantees bandwidth and eliminates jitter (timing variations), making it ideal for voice. However, it's inefficient for bursty data traffic because the circuit remains reserved even during silence in conversations, wasting capacity that could serve other users.
The PSTN infrastructure consists of millions of kilometers of copper wire, fiber optic cables, and switching equipment connecting homes, businesses, and mobile networks worldwide. At its peak in the early 2000s, PSTN served over 1.3 billion landline subscribers globally. Today, as Voice over IP (VoIP) and mobile networks dominate, many countries are gradually decommissioning their PSTN infrastructure in a process called "PSTN switch-off," with complete shutdowns planned between 2025-2030 in markets like the UK, Germany, and Switzerland.
Dedicated path for entire call duration
Guaranteed bandwidth and low latency
Worldwide interconnected network
| Component | Description | Function |
|---|---|---|
| Subscriber Line | Copper wire pair connecting home/business to CO | Last-mile connectivity, analog voice transmission |
| Central Office (CO) | Local telephone exchange serving area subscribers | Call switching, routing, subscriber services |
| Tandem Office | Interconnects multiple central offices | Inter-office call routing within region |
| Toll Center | Handles long-distance calls | Inter-city/state call routing |
| Trunk Lines | High-capacity links between exchanges | Aggregate multiple calls over shared media |
| Signaling System 7 (SS7) | Out-of-band signaling network | Call setup, routing, billing, advanced features |
The PSTN hierarchy follows a Class system where Class 5 switches (local exchanges) serve end users directly, while Class 4, 3, 2, and 1 switches handle progressively larger geographic areas. When you make a call, it first goes to your local Class 5 exchange. If the destination is within the same exchange, the call is switched locally. For long-distance calls, the call is routed up the hierarchy to a tandem office or toll center, then back down to the destination's local exchange.
SS7 (Signaling System 7) revolutionized PSTN by separating signaling from voice traffic. Instead of using the same circuit for both call setup information and the actual voice, SS7 uses a dedicated packet-switched network to exchange signaling messages. This enables features like Caller ID, call forwarding, toll-free numbers, and number portability. SS7 messages contain information about the calling party, called party, call type, and routing instructions, allowing switches to establish connections intelligently.
The local loop (also called "last mile") connecting homes to the central office typically uses twisted-pair copper wires. These analog connections carry voice frequencies (300-3400 Hz) and can span several kilometers. Digital Subscriber Line (DSL) technology later repurposed these same copper pairs to deliver broadband internet by using higher frequency ranges above the voice band, allowing simultaneous voice and data transmission.
| Component | Description | Function |
|---|---|---|
| Subscriber Line | Copper wire pair connecting home/business to CO | Last-mile connectivity, analog voice transmission |
| Central Office (CO) | Local telephone exchange serving area subscribers | Call switching, routing, subscriber services |
| Tandem Office | Interconnects multiple central offices | Inter-office call routing within region |
| Toll Center | Handles long-distance calls | Inter-city/state call routing |
| Trunk Lines | High-capacity links between exchanges | Aggregate multiple calls over shared media |
| Signaling System 7 (SS7) | Out-of-band signaling network | Call setup, routing, billing, advanced features |
The PSTN hierarchy follows a Class system where Class 5 switches (local exchanges) serve end users directly, while Class 4, 3, 2, and 1 switches handle progressively larger geographic areas. When you make a call, it first goes to your local Class 5 exchange. If the destination is within the same exchange, the call is switched locally. For long-distance calls, the call is routed up the hierarchy to a tandem office or toll center, then back down to the destination's local exchange.
SS7 (Signaling System 7) revolutionized PSTN by separating signaling from voice traffic. Instead of using the same circuit for both call setup information and the actual voice, SS7 uses a dedicated packet-switched network to exchange signaling messages. This enables features like Caller ID, call forwarding, toll-free numbers, and number portability. SS7 messages contain information about the calling party, called party, call type, and routing instructions, allowing switches to establish connections intelligently.
The local loop (also called "last mile") connecting homes to the central office typically uses twisted-pair copper wires. These analog connections carry voice frequencies (300-3400 Hz) and can span several kilometers. Digital Subscriber Line (DSL) technology later repurposed these same copper pairs to deliver broadband internet by using higher frequency ranges above the voice band, allowing simultaneous voice and data transmission.
The evolution of mobile networks follows a clear generational pattern, with each "G" representing a fundamental technological leap rather than incremental improvement. Each generation introduced new radio access technologies, modulation schemes, network architectures, and use cases that defined how people communicate and access information. The progression from 1G to 5G took approximately 40 years, with each generation having a lifecycle of about 10-15 years before being superseded.
What drives these generational shifts? Primarily, it's the insatiable demand for higher data speeds, lower latency, greater device density, and improved energy efficiency. Each generation required developing new international standards (like 3GPP specifications), allocating new radio spectrum, upgrading base station equipment, and deploying new core network infrastructure. Operators typically maintain multiple generations simultaneously during transition periods—today's networks often support 3G, 4G, and 5G concurrently, with devices automatically selecting the best available network.
| Generation | Technology | Data Speed | Key Features | Primary Use |
|---|---|---|---|---|
| 1G 1980s |
AMPS, NMT Analog |
~2.4 kbps |
• Analog voice only • Large cell coverage • No encryption |
Voice calls |
| 2G 1991 |
GSM, CDMA Digital |
64 kbps |
• Digital voice • SMS & MMS • SIM cards • Basic encryption |
Voice + Text |
| 3G 2001 |
UMTS, HSPA W-CDMA |
2-42 Mbps |
• Mobile internet • Video calls • Mobile TV • GPS integration |
Mobile Data |
| 4G LTE 2009 |
LTE, LTE-A OFDMA |
100-1000 Mbps |
• All-IP network • HD video streaming • VoLTE • Low latency (~50ms) |
Broadband Mobile |
| 5G NR 2019+ |
5G NR mmWave, Sub-6 |
1-20 Gbps |
• Massive IoT • Ultra-low latency (~1ms) • Network slicing • Edge computing |
IoT, AR/VR, Industry 4.0 |
1G to 2G: Analog to Digital
The transition from 1G to 2G wasn't just about digitization—it fundamentally changed what phones could do. Digital encoding enabled SMS text messaging, which became unexpectedly popular. GSM's SIM card architecture allowed users to switch phones easily, spurring device competition and innovation. Digital signals were more resistant to interference and eavesdropping, improving both quality and security.
2G to 3G: Voice to Data
3G introduced mobile internet, enabling mobile email, web browsing, and early app stores. The iPhone's 2007 launch on 3G networks catalyzed the smartphone revolution. HSPA+ (High-Speed Packet Access) evolved 3G to deliver up to 42 Mbps, competitive with early 4G. Video calling became practical, though it never achieved mass adoption due to user behavior preferences.
3G to 4G: Mobile Broadband
4G LTE represented a complete architectural overhaul. Unlike previous generations that were primarily enhancements of circuit-switched networks, 4G was all-IP from the ground up. Voice calls used VoLTE (Voice over LTE), treating voice as just another IP application. This enabled HD video streaming, mobile gaming, and real-time applications. LTE's OFDMA (Orthogonal Frequency Division Multiple Access) dramatically improved spectral efficiency, allowing more users per cell tower.
4G to 5G: Beyond Human Communication
5G isn't primarily about making smartphones faster—it's about connecting everything. The three use case categories (eMBB, URLLC, mMTC) target distinct applications: eMBB serves traditional smartphone users with gigabit speeds, URLLC enables mission-critical applications requiring sub-millisecond latency (autonomous vehicles, industrial robotics, remote surgery), and mMTC connects millions of IoT sensors per square kilometer for smart cities and industrial monitoring. Network slicing allows operators to create virtualized networks with specific characteristics for each use case on shared physical infrastructure.
Cellular networks divide geographic areas into cells, each served by a base station. This architecture enables frequency reuse, increasing network capacity while managing interference.
The cellular concept, pioneered by Bell Labs in the 1960s-70s, solved a fundamental problem: limited radio spectrum. Since radio frequencies are finite and valuable, how could networks serve millions of users? The solution was frequency reuse—using the same frequencies in geographically separated cells, with cell towers transmitting at lower power to prevent interference with distant cells using the same frequencies.
Cell towers don't actually create circular cells. The hexagonal cell pattern often shown in diagrams is an idealization. Real-world cells are irregular shapes determined by terrain, buildings, tower height, antenna configuration, and transmit power. Urban areas have small cells (macrocells 1-2 km radius, microcells 200-500m, picocells 50-200m, femtocells 10-50m) providing high capacity, while rural areas use larger macrocells (up to 35 km radius) for coverage over sparse populations.
Modern cellular networks use a three-layer architecture: Radio Access Network (RAN) consisting of base stations (eNodeB in 4G, gNodeB in 5G) and antennas that communicate with mobile devices over radio frequencies; Backhaul/Transport Network using fiber optics or microwave links to connect base stations to the core; and Core Network (EPC in 4G, 5GC in 5G) handling authentication, session management, mobility, policy enforcement, and connectivity to external networks like the Internet.
The evolution from 4G EPC (Evolved Packet Core) to 5G Core represents a fundamental architectural transformation. While 4G's EPC was built on traditional telecom principles with monolithic network functions, 5G Core embraces cloud-native architecture using microservices, containerization, and service-based interfaces.
Service-Based Architecture (SBA): In 5G Core, network functions expose their services through standardized APIs, allowing flexible composition and orchestration. This contrasts with 4G's point-to-point interfaces between network elements. SBA enables rapid service deployment, third-party integration, and network function reuse across different services.
Control and User Plane Separation (CUPS): 5G strictly separates control plane (signaling) from user plane (actual data forwarding), allowing independent scaling. During peak traffic, operators can add more UPF (User Plane Function) instances without touching control plane functions. This improves efficiency and enables edge computing deployment where UPF instances run close to users for ultra-low latency.
Network Slicing: Perhaps 5G Core's most revolutionary feature, network slicing creates multiple virtual networks on shared physical infrastructure. Each slice has specific characteristics (bandwidth, latency, reliability) tailored for different use cases. An autonomous vehicle slice prioritizes ultra-low latency, while an IoT slice optimizes for massive device connectivity and power efficiency. The NSSF (Network Slice Selection Function) directs devices to appropriate slices during registration.
Different mobile generations use different frequency bands. Lower frequencies offer better coverage and penetration, while higher frequencies provide greater bandwidth and speed.
Radio spectrum is governed by physics: lower frequencies (below 1 GHz) have longer wavelengths that diffract around obstacles and penetrate buildings effectively, providing wide coverage areas but limited bandwidth. Higher frequencies (above 6 GHz) have shorter wavelengths carrying more data but requiring line-of-sight and struggling with obstacles. This creates a fundamental tradeoff between coverage and capacity.
Spectrum Licensing and Auctions: Governments allocate spectrum through regulatory bodies (FCC in USA, Ofcom in UK, etc.) via competitive auctions where operators bid billions of dollars for frequency licenses. The 2021 US C-band auction raised $81 billion, the largest in FCC history. Operators must carefully balance spectrum portfolio across low, mid, and high bands to deliver both coverage and capacity.
Licensed vs Unlicensed Spectrum: Most cellular networks use licensed spectrum with exclusive rights, preventing interference. However, technologies like LTE-U and 5G NR-U also leverage unlicensed spectrum (like WiFi's 2.4/5 GHz bands) to boost capacity. This coexistence requires sophisticated interference management through techniques like Listen Before Talk (LBT), where equipment senses the channel before transmitting.
Frequency: 600-900 MHz
Coverage: Wide area (10-100 km)
Speed: 50-250 Mbps
Use: Rural, IoT, nationwide coverage
Frequency: 3.3-4.9 GHz
Coverage: Urban (1-5 km)
Speed: 100-900 Mbps
Use: Cities, capacity, balance
Frequency: 24-39 GHz
Coverage: Micro (100-500 m)
Speed: 1-10 Gbps
Use: Stadiums, venues, hotspots
Carrier Aggregation: Modern networks combine multiple frequency bands simultaneously to increase throughput. A 5G device might aggregate 100 MHz of C-band spectrum with 20 MHz of low-band spectrum, achieving higher speeds while maintaining reliability. Advanced implementations support up to 5 carriers across different bands, with intra-band (same band, different channels) and inter-band (different bands) aggregation.
Dynamic Spectrum Sharing (DSS): Rather than dedicating spectrum exclusively to 4G or 5G, DSS allows real-time sharing of the same frequency band between both technologies. The network dynamically allocates resources based on demand - if more users are on 5G devices, more spectrum is allocated to 5G, and vice versa. This facilitates smooth 5G migration without requiring dedicated spectrum.
Massive MIMO and Beamforming: Traditional base stations use 2-8 antennas broadcasting in all directions. Massive MIMO employs 64-256 antennas creating focused beams toward individual users through beamforming. This dramatically increases capacity (10x+) and improves signal quality by reducing interference. In mmWave deployments, beamforming is essential because high-frequency signals attenuate rapidly - narrow beams concentrate energy toward receivers.
Full Duplex and TDD/FDD: Cellular systems use either Frequency Division Duplex (FDD) with separate frequency bands for uplink and downlink, or Time Division Duplex (TDD) alternating transmit/receive on the same frequency. 5G favors TDD because it allows flexible uplink/downlink ratios and works well with massive MIMO. Experimental full-duplex systems transmit and receive simultaneously on the same frequency, potentially doubling spectral efficiency, but require sophisticated self-interference cancellation.
Handoff is the process of transferring an ongoing call or data session from one cell to another as a mobile device moves. This ensures seamless connectivity without dropping the connection.
Handoff is one of the most complex and critical procedures in cellular networks. Without it, calls would drop every time you drove past a cell tower or walked between buildings. The network must balance competing objectives: handoff too early and you waste network resources with unnecessary transfers; handoff too late and calls drop when signal quality deteriorates beyond recovery.
Measurement and Decision: Mobile devices continuously measure signal strength and quality from their serving cell and neighboring cells, reporting these measurements to the network. The network analyzes these reports using handoff algorithms considering signal strength (RSRP - Reference Signal Received Power), signal quality (RSRQ - Reference Signal Received Quality), and signal-to-interference ratio. When measurements indicate a neighbor cell provides better service, the network initiates handoff.
Hard vs Soft Handoff: GSM and LTE use "hard handoff" (break-before-make) where the connection to the old cell terminates before establishing connection to the new cell, causing brief interruptions (typically 50-100ms). CDMA-based networks (including some 3G systems) use "soft handoff" (make-before-break) where devices maintain simultaneous connections to multiple cells during transition, improving reliability but requiring more network resources. 5G introduces sophisticated handoff procedures supporting ultra-low latency applications with sub-millisecond interruption times.
Used in: 2G/3G CDMA, older systems
Process: Break-before-make
Used in: 4G LTE, 5G
Process: Make-before-break
Roaming allows mobile users to access services outside their home network by using visited networks. This involves authentication, billing, and inter-operator agreements.
Roaming is essential for mobile telecommunications, enabling the global connectivity we take for granted. When you travel internationally, your phone seamlessly connects to foreign networks without requiring new SIM cards or contracts. Behind this convenience lies complex technical and business arrangements between hundreds of network operators worldwide.
How Roaming Works: When your device powers on in a foreign country, it scans for available networks and selects one based on your home operator's roaming agreements. The visited network queries your home network's HLR/HSS (subscriber database) through SS7 or Diameter signaling to authenticate your identity and retrieve your service profile. Once authenticated, a temporary record is created in the visited network's VLR, and you can make calls, send messages, and use data.
Roaming Agreements and Wholesale Rates: Operators establish bilateral or multilateral roaming agreements defining technical parameters and commercial terms. When you use data while roaming, your home operator pays the visited operator a wholesale rate (typically per MB or per minute for calls), then charges you retail roaming rates. This multi-layered pricing explains why roaming was historically expensive, though EU regulations and competitive pressure have dramatically reduced costs in many markets.
Roaming Steering and Optimization: Your device doesn't necessarily connect to the network with the strongest signal. Home operators can steer devices toward preferred roaming partners offering better wholesale rates or quality. This steering occurs through preferred network lists in the SIM card and over-the-air updates. Advanced systems analyze real-time quality metrics and costs to optimize roaming partner selection automatically.
| Roaming Type | Description | Authentication | Use Case |
|---|---|---|---|
| National Roaming | Within same country, different operator | Home HLR/HSS validates via visited network | Rural coverage extension, MVNO services |
| International Roaming | Across countries, foreign operators | IMSI-based authentication through SS7/Diameter | Travel, business trips abroad |
| Permanent Roaming | Long-term use outside home network | Same as international, special billing | Long-term stay, expatriates |
| Data Roaming | Internet access while roaming | APN configuration + standard auth | Mobile data abroad, higher rates |
Network slicing is a revolutionary 5G feature that allows operators to create multiple virtual networks on a single physical infrastructure. Each slice is customized for specific use cases with dedicated resources, SLAs, and performance characteristics.
| Feature | 4G LTE | 5G NR | Improvement |
|---|---|---|---|
| Peak Data Rate | 1 Gbps | 20 Gbps | 20x faster |
| Latency | 50ms | 1ms | 50x lower |
| Device Density | 100K/km² | 1M/km² | 10x more devices |
| Mobility | 350 km/h | 500 km/h | 43% faster |
| Energy Efficiency | Baseline | 90% reduction | 10x more efficient |
| Spectrum | Sub-6 GHz | Sub-6 + mmWave (24-100 GHz) | Much wider range |
| Network Slicing | Not available | Yes (virtual networks) | New capability |
| Edge Computing | Limited | Native MEC support | Distributed processing |
Multiple-Input Multiple-Output
Millimeter Wave Spectrum
Multi-Access Edge Computing
DSS Technology
Performance Targets:
Key Innovations:
In the next topic, we'll explore IP Networking and Data Center Architecture. You'll learn about: