Zum Inhalt springen

Cellular Networks: From 1G to 5G

Zusammenfassung

Cellular networks are the radio infrastructure that made the mobile phone possible — and then made the smartphone, mobile internet, and the always-connected world inevitable. The defining idea is deceptively simple: divide a service area into “cells,” each served by a low-power base station, and reuse the same scarce radio frequencies in non-adjacent cells. That insight, formalized at Bell Labs in 1947 and finally made practical in the 1970s and 1980s, broke the fundamental capacity ceiling of earlier mobile radio. Over five generations — analog voice (1G), digital voice and texting (2G), mobile data (3G), mobile broadband (4G/LTE), and ultra-low-latency wide-bandwidth networks (5G) — cellular technology went from a luxury for a few thousand car-phone subscribers to a system carrying voice and data for more than 8 billion connections worldwide. This article traces the radio networks themselves, distinct from the devices and operating systems that ride on top of them.

The Cellular Idea

Mobile radio telephony predates the cellular concept by decades. Early systems — such as AT&T’s Mobile Telephone Service (MTS), launched in 1946 in St. Louis — used a single powerful transmitter covering an entire city. The problem was capacity: a city-wide transmitter could only support a handful of simultaneous calls, because every conversation consumed one of a small number of channels across the whole coverage area. By the 1970s, waiting lists for mobile service in major American cities ran for years, and the number of subscribers was capped in the low thousands.

The solution had been described in an internal Bell Labs memo by Douglas H. Ring and W. Rae Young in 1947: divide the territory into many small cells, each with its own low-power transmitter, and reuse the same frequencies in cells far enough apart that they would not interfere. Instead of a few channels shared across a city, the same channels could be used over and over. As demand grew, cells could be “split” into smaller cells, multiplying capacity further.

The idea was decades ahead of the electronics needed to realize it. The crucial unsolved problem was handoff (or handover): as a moving vehicle crossed from one cell to the next, the call had to be transferred to the new base station’s frequency seamlessly, in real time, under computer control. Amos E. Joel Jr. of Bell Labs developed the handoff switching system, and the full architecture was laid out in a 1970s Bell Labs proposal led by engineers including Richard H. Frenkiel and Joel Engel, who later received the National Medal of Technology for the work. The first commercial deployment, however, happened not in the United States but in Japan: NTT launched a commercial cellular network in Tokyo in 1979. The Nordic countries followed with the NMT (Nordic Mobile Telephone) system in 1981, and the U.S. AMPS (Advanced Mobile Phone System) network launched commercially in Chicago in 1983.

1G: Analog Voice

The first generation of cellular was analog. Voice was transmitted as a frequency-modulated radio signal, much like FM radio, using FDMA (Frequency Division Multiple Access) — each call got its own frequency channel. The dominant systems were AMPS in North America, NMT in the Nordics, and TACS in the UK.

1G worked, and it created the first mass mobile-phone market, but its limitations were severe. Analog signals were trivial to intercept with a scanner — eavesdropping on calls was a notorious problem. There was no encryption and weak authentication, so “cloning” a phone’s identity to steal service was rampant. Capacity remained limited, voice quality degraded gracefully but noticeably with distance and interference, and the systems of different countries were incompatible. The handsets were large and power-hungry; the iconic Motorola DynaTAC 8000X (1983), the first commercially available handheld cellular phone, weighed about 790 grams and offered roughly 30 minutes of talk time.

2G: Going Digital — and the GSM Standard

The second generation, arriving in the early 1990s, was digital. Voice was now sampled, compressed, and transmitted as bits. This brought encryption, far better spectral efficiency, and — almost as an afterthought — text messaging.

The defining 2G standard was GSM (Global System for Mobile Communications), developed under the auspices of European telecom administrations (the CEPT, later ETSI). GSM’s first commercial network launched in Finland in 1991 (Radiolinja). GSM used TDMA (Time Division Multiple Access), dividing each frequency into time slots so multiple calls could share one channel. Crucially, GSM was a true international standard: a GSM phone could roam across borders, and the SIM card (Subscriber Identity Module) decoupled the subscriber’s identity from the handset — a SIM could be moved between phones. GSM became the most widely adopted cellular standard in history, deployed in over 200 countries.

In North America, a competing approach gained ground: CDMA (Code Division Multiple Access), commercialized by Qualcomm as IS-95 (cdmaOne). Instead of dividing by frequency or time, CDMA let all users transmit across the whole band simultaneously, separating signals by unique mathematical codes — a spread-spectrum technique. CDMA offered higher capacity but split the global market into incompatible GSM and CDMA camps, a division that persisted for two decades.

2G’s quiet revolution was SMS (Short Message Service). Designed as a minor signaling feature with a 160-character limit — a constraint reportedly chosen by engineer Friedhelm Hillebrand after analyzing the typical length of postcards and Telex messages — the first SMS (“Merry Christmas”) was sent in December 1992. Text messaging became a cultural and economic phenomenon, generating enormous revenue for carriers and reshaping how people communicated. Data, via GPRS and later EDGE (sometimes called “2.5G”), arrived as a slow but real packet-data overlay.

3G: The Mobile Internet

The third generation, standardized under the ITU’s IMT-2000 framework and arriving commercially around 2001 (NTT DoCoMo’s FOMA in Japan was first), was built for data. The dominant 3G standard, UMTS (Universal Mobile Telecommunications System), used W-CDMA (Wideband CDMA) — vindicating Qualcomm’s CDMA bet by making it the radio foundation even of the GSM lineage’s successor. Qualcomm’s patent position made it a beneficiary of nearly every 3G phone sold.

3G delivered data rates that made mobile web browsing, email, and eventually app stores and video genuinely usable. Successive enhancements — HSPA (High Speed Packet Access) and HSPA+ — pushed speeds higher. The arrival of the iPhone (2007) and Android (2008) on 3G networks turned mobile data from a niche feature into the primary use of a phone, straining networks that had been provisioned mainly for voice.

4G / LTE: Mobile Broadband

The fourth generation abandoned the old circuit-switched voice architecture entirely. LTE (Long-Term Evolution), standardized by the 3GPP and first deployed commercially by TeliaSonera in Stockholm and Oslo in 2009, is all-IP: every payload, including voice, is carried as data packets. Voice itself moved to VoLTE (Voice over LTE), treating calls as just another internet application.

LTE’s radio used OFDMA (Orthogonal Frequency Division Multiple Access) and MIMO (Multiple-Input Multiple-Output) antennas — using multiple antennas at both ends to multiply throughput. Real-world speeds of tens to hundreds of megabits per second made smartphone video streaming, mapping, ride-hailing, and the entire mobile-app economy practical. (Early “4G” marketing was applied loosely; the ITU’s strict IMT-Advanced criteria were really met by LTE-Advanced, but the “4G” label stuck to LTE in common usage.)

5G: Bandwidth, Latency, and Hype

The fifth generation, with standards finalized by 3GPP from 2018 (Release 15) onward and first commercial launches in 2019 (South Korea and the U.S.), promised three things: much higher peak bandwidth (eMBB), very low latency (URLLC, targeting around 1 millisecond for applications like industrial automation and remote control), and massive device density (mMTC, for IoT). 5G introduced the use of millimeter-wave (mmWave) spectrum — very high frequencies offering enormous bandwidth but extremely short range and poor wall penetration — alongside more practical mid-band (“sub-6 GHz”) deployments that carry most real-world 5G traffic.

5G also brought network slicing (partitioning one physical network into multiple virtual networks tuned for different uses) and massive MIMO (base stations with dozens or hundreds of antenna elements steering focused beams toward users). The technology’s rollout was accompanied by considerable hype, geopolitical conflict over equipment vendors (notably the security debate around Huawei), and — bizarrely — conspiracy theories falsely linking 5G to COVID-19 that led to arson attacks on cell towers.

Dead End: WAP and the “Walled Garden” Mobile Web

Not every step forward worked. Around 1999–2000, before real mobile broadband existed, the industry promoted WAP (Wireless Application Protocol) as the way to bring the internet to phones. WAP pages were written in a stripped-down markup language (WML) and viewed on tiny monochrome screens over painfully slow 2G data links. It was slow, expensive (often billed by the minute), clumsy to navigate, and confined to a handful of carrier-approved “walled garden” sites. WAP became a byword for overhyped disappointment in much of the world.

The instructive counterexample was Japan’s i-mode, launched by NTT DoCoMo in 1999, which succeeded spectacularly with a similar concept — because DoCoMo got the business model right (an open content ecosystem with built-in micropayments and revenue sharing for developers) rather than because the technology was fundamentally different. WAP’s failure showed that mobile internet would not arrive until two things matured together: fast packet data (3G) and capable devices with real browsers (the iPhone). When they did, the walled gardens were swept away by the open mobile web and native app stores.

📚 Sources