3G and 4G Mobile Networks

In the early 2000s, the telecommunications industry faced a critical challenge: how to deliver high-speed data services to mobile devices in an era when 2G networks—designed primarily for voice with painfully slow data rates of 9.6 to 14.4 kbps—simply could not meet growing user expectations.

The answer was Third Generation (3G) mobile networks—a revolutionary leap that redefined what mobile connectivity could achieve. 3G didn't just incrementally improve upon 2G; it fundamentally transformed mobile networks from voice-centric systems into multimedia-capable broadband platforms supporting video calling, mobile internet browsing, and streaming services.

Understanding 3G is essential not just for historical perspective but because many of its architectural innovations—such as code division multiple access, soft handoffs, and packet-switched core networks—remain foundational concepts in modern 4G and 5G systems. The 3G era established the engineering patterns that enabled today's mobile broadband experience.

To appreciate 3G's significance, we must first understand why 2G (GSM/CDMA IS-95) systems proved inadequate for the emerging mobile data era:

Fundamental 2G Limitations:

1. Extreme Data Rate Constraints: GSM's circuit-switched data provided only 9.6-14.4 kbps—slower than dial-up modems of the era. Loading a simple web page took minutes.

2. Circuit-Switched Architecture: Every data connection occupied an entire circuit continuously, even during idle periods. This was monumentally inefficient for bursty internet traffic where devices alternate between active transmission and waiting.

3. Voice-Centric Design: 2G systems were optimized for voice with fixed 13 kbps voice codecs and rigid time-slot allocations. Data was an afterthought, retrofitted onto voice infrastructure.

4. Limited Spectrum Efficiency: TDMA (Time Division Multiple Access) and FDMA (Frequency Division Multiple Access) used in 2G provided limited capacity and suffered from hard capacity limits due to inter-cell interference.

5. Poor Roaming and Fragmentation: Different 2G standards (GSM in Europe, CDMA/TDMA in North America) created fragmented global coverage, complicating international roaming.

The 2.5G Bridge Attempts:

Technologies like GPRS (General Packet Radio Service) and EDGE (Enhanced Data rates for GSM Evolution) attempted to retrofit packet-switching onto 2G GSM infrastructure. While they represented genuine improvements—GPRS introduced packet-switched data and EDGE improved modulation efficiency—they were fundamentally band-aids on an architecture not designed for data.

GPRS Architecture Insight:

• Packet-Switched Overlay: GPRS added new network nodes (SGSN, GGSN) to handle packet data alongside the existing circuit-switched voice infrastructure
• Shared Radio Resources: Multiple users shared time slots using statistical multiplexing, improving efficiency
• Always-On Capability: Devices could maintain IP connectivity without occupying circuits continuously

However, GPRS/EDGE shared radio resources with voice, creating contention. During busy periods, data performance degraded significantly. The world needed a clean-slate approach designed from the ground up for mobile broadband.

Code Division Multiple Access (CDMA) represents the technological heart of 3G systems. Unlike 2G's time-slot (TDMA) or frequency-slot (FDMA) divisions, CDMA allows all users to transmit simultaneously over the entire frequency band by distinguishing signals through unique orthogonal codes.

Conceptual Understanding:

Imagine a cocktail party where everyone speaks simultaneously. In TDMA, people take turns speaking. In FDMA, each person speaks at a different pitch (frequency). In CDMA, everyone speaks simultaneously in the same frequency range, but each pair of conversants uses a unique language that only they understand. Despite the apparent chaos, each listener can extract their conversation by "filtering out" all other languages.

Mathematical Foundation:

In CDMA, each user is assigned a unique spreading code—a pseudo-random sequence of chips that spreads the signal across a wide frequency band. At the receiver, the same spreading code is used to despread the intended signal while reducing interference from other users to noise-like levels.

Spreading Process:

Original data bit:       1 (duration: T_b)
Spreading code:          +1, -1, +1, +1, -1, -1, +1, -1 (8 chips, duration: T_c each)
Spread signal:           +1, -1, +1, +1, -1, -1, +1, -1 (if bit = 1)
                         -1, +1, -1, -1, +1, +1, -1, +1 (if bit = 0)

The spreading factor (or processing gain) = T_b / T_c determines interference suppression. A spreading factor of 128 means each data bit becomes 128 chips, providing approximately 21 dB of processing gain against interference.

Spreading Code Types:

3G systems employ different code families for different purposes:

1. Walsh-Hadamard Codes (Orthogonal): Used for channelization within a single cell on the downlink. These codes are perfectly orthogonal—their cross-correlation is zero—allowing interference-free separation of user channels within a cell.

2. Pseudo-Noise (PN) Sequences: Long PN sequences distinguish cells from each other (scrambling codes). They have low but non-zero cross-correlation, managing inter-cell interference statistically.

3. Gold Codes: A family of codes with bounded cross-correlation properties, often used for scrambling in UMTS.

Power Control Criticality:

CDMA systems face the near-far problem: a mobile device close to a base station could overwhelm signals from distant devices if all transmitted at equal power. 3G CDMA implements tight closed-loop power control adjusting transmit power up to 1,500 times per second (every 0.67 ms in CDMA2000) to ensure all signals arrive at the base station at similar power levels.

This power control is mission-critical: a 1 dB deviation from optimal can translate to measurable capacity loss. The precision required is remarkable—power control commands adjust transmitter output by fractions of a dB based on real-time channel measurements.

Universal Mobile Telecommunications System (UMTS) emerged as the dominant 3G standard globally, developed by the 3rd Generation Partnership Project (3GPP) as the evolution of GSM. Its radio access technology, Wideband CDMA (WCDMA), represents a clean-break redesign optimized for broadband data.

WCDMA Key Specifications:

UMTS Network Architecture:

UMTS introduced a new architecture with clearly separated domains:

1. User Equipment (UE): The mobile device containing:

   • Mobile Equipment (ME): The physical device
   • USIM (UMTS Subscriber Identity Module): Enhanced SIM with stronger security

2. UTRAN (UMTS Terrestrial Radio Access Network):

   • Node B: The base station handling radio transmission/reception (analogous to GSM's BTS)
   • RNC (Radio Network Controller): Manages multiple Node Bs, handling handovers, radio resource management, and flow control (analogous to GSM's BSC but with expanded responsibilities)

3. Core Network: Dual-domain architecture

   • Circuit-Switched (CS) Domain: MSC, GMSC for voice calls—inherited from GSM
   • Packet-Switched (PS) Domain: SGSN, GGSN for data—evolved from GPRS

WCDMA Channel Structure:

WCDMA employs a sophisticated hierarchy of channels:

Logical Channels (what is transferred):

• Broadcast Control Channel (BCCH): System information
• Paging Control Channel (PCCH): Paging messages
• Dedicated Control Channel (DCCH): Signaling for connected users
• Dedicated Traffic Channel (DTCH): User data

Transport Channels (how it's transferred):

• Broadcast Channel (BCH): Carries BCCH
• Paging Channel (PCH): Carries PCCH
• Dedicated Channel (DCH): Bidirectional, supports soft handoff
• High-Speed Downlink Shared Channel (HS-DSCH): Added in HSDPA

Physical Channels (where it goes):

• Primary Common Pilot Channel (P-CPICH): Cell reference signal
• Primary/Secondary Synchronization Channels (P-SCH, S-SCH): Timing acquisition
• Common Pilot Channel (CPICH): Channel estimation
• Dedicated Physical Data/Control Channels (DPDCH, DPCCH): User data and control

Soft Handoff in WCDMA:

Unlike GSM's "break-before-make" hard handoffs, WCDMA implements soft handoff where the mobile communicates with multiple cells simultaneously within an "Active Set." On the downlink, the mobile combines signals from multiple cells using maximal-ratio combining. On the uplink, multiple RNCs receive the signal and the best frame is selected (selection combining). This provides:

• Macro-diversity gain against fading
• Seamless user experience during mobility
• Reduced ping-pong handovers at cell boundaries

CDMA2000 emerged as the 3G evolution of IS-95 (cdmaOne), developed by 3GPP2 primarily for North American and Asian operators who had deployed IS-95 networks. While UMTS required greenfield spectrum deployment, CDMA2000 was designed for backward compatibility with existing IS-95 spectrum and infrastructure.

CDMA2000 Evolution Phases:

1. CDMA2000 1xRTT (1x Radio Transmission Technology):

   • First phase, using single 1.25 MHz carrier (like IS-95)
   • Peak data rate: 153 kbps
   • Doubled voice capacity over IS-95 through improved algorithms
   • Backward compatible with existing IS-95 handsets for voice

2. CDMA2000 1xEV-DO (Evolution-Data Optimized):

   • Data-only carrier overlay; separate from voice channels
   • Rev. 0: 2.4 Mbps peak downlink
   • Rev. A: 3.1 Mbps downlink, 1.8 Mbps uplink
   • Rev. B: Multi-carrier aggregation up to 14.7 Mbps
   • Time-shared downlink optimized for bursty data traffic

3. CDMA2000 1xEV-DV (Evolution-Data/Voice):

   • Integrated voice and data on single carrier
   • Never widely deployed; EV-DO became the preferred path

1xEV-DO: A Paradigm Shift in Radio Design:

1xEV-DO introduced revolutionary concepts later adopted by 4G:

Time-Shared Downlink: Rather than giving each user a dedicated channel at fixed rate, 1xEV-DO transmits to users one at a time at maximum possible rate for their channel conditions, then quickly switches. This approach:

• Maximizes total throughput (sends more when channels are good)
• Trades per-user rate control for scheduler-managed fairness
• Achieves multi-user diversity gain

Proportional Fair Scheduling: The scheduler balances:

• Users with good instantaneous channels (for throughput)
• Users waiting longest relative to their average rate (for fairness)

Adaptive Modulation and Coding (AMC): Data rates adapt rapidly based on feedback:

• Good channel: 16-QAM with rate 3/4 coding → maximum throughput
• Poor channel: QPSK with rate 1/4 coding → robust but slower
• Decisions made every 1.67 ms slot

These innovations made 1xEV-DO remarkably efficient for data, achieving near-Shannon-limit spectral efficiency. The techniques directly influenced 4G LTE's downlink design.

As mobile data consumption surged beyond initial projections, 3GPP introduced HSPA (High Speed Packet Access)—a family of enhancements that dramatically boosted UMTS performance without requiring new spectrum or fundamental architecture changes.

HSDPA (High Speed Downlink Packet Access) - 3GPP Release 5:

HSDPA transformed UMTS downlink performance through innovations borrowed from CDMA2000 EV-DO:

HSDPA Peak Rates:

• Category 1-4: 1.8 Mbps
• Category 5-6: 3.6 Mbps
• Category 7-8: 7.2 Mbps
• Category 9-10: 10.1 Mbps
• Category 11-12: 14.4 Mbps (16-QAM, 15 codes)

HSUPA (High Speed Uplink Packet Access) - 3GPP Release 6:

HSUPA enhanced the uplink with similar principles adapted for the asymmetric uplink environment:

• Enhanced Dedicated Channel (E-DCH): New uplink transport channel optimized for bursty data
• Shorter TTI: 2 ms or 10 ms options for latency/coverage tradeoff
• Node B Scheduling: Fast grant-based scheduling reduces latency
• Soft Handoff Support: Unlike HSDPA, supports macro-diversity
• Multi-code Transmission: Up to 4 channelization codes per user
• Peak Rate: Up to 5.76 Mbps

HSPA+ (Evolved HSPA) - 3GPP Releases 7-11:

Continued evolution pushed 3G to remarkable performance:

These enhancements extended UMTS lifespan significantly, allowing operators to maximize 3G investments while preparing 4G deployments.

3G Spectrum Bands:

3G systems operate across multiple frequency bands, allocated through ITU coordination and national regulatory processes:

Primary IMT-2000 Bands:

Spectrum Characteristics:

1. Lower Bands (850/900 MHz):

   • Superior propagation, building penetration
   • Larger cell coverage areas—fewer sites needed
   • Limited bandwidth availability (often refarmed from 2G)
   • Better for rural coverage extension

2. Higher Bands (1900/2100 MHz):

   • More spectrum typically available
   • Better capacity in urban areas
   • Smaller cell radius requires more sites
   • Primary bands for initial 3G launches

Global Deployment Timeline:

3G deployment occurred in waves, determined by spectrum availability, regulatory frameworks, and market conditions:

Early Adopters (2001-2003):

• Japan (2001): NTT DoCoMo launched world's first commercial 3G (WCDMA) in October 2001
• Korea (2002): CDMA2000 deployment, later supplemented by WCDMA
• Europe (2003): Major operators like Hutchison (3), Vodafone, Orange launched UMTS

Main Wave (2004-2007):

• North America: Verizon, Sprint (CDMA2000 EV-DO); AT&T, T-Mobile (UMTS)
• Asia-Pacific: Widespread UMTS deployment in China, India, Southeast Asia
• Latin America: Mixed CDMA2000 and UMTS deployments

Optimization Phase (2008-2012):

• HSPA/HSPA+ rollout dramatically improved performance
• Spectrum refarming: 900 MHz UMTS for coverage, 2100 MHz for capacity
• Continued expansion in emerging markets

Sunset Planning (2015-Present):

• Operators began 3G shutdown announcements
• 2G often retained longer for IoT/M2M
• Spectrum refarmed to 4G/5G
• Many 3G networks scheduled for 2022-2025 shutdown

3G represented a paradigm shift in mobile communications, transforming phones from voice devices into mobile broadband terminals. Let's consolidate the key technological contributions:

Limitations That Drove 4G Development:

Despite HSPA+ achieving impressive rates, 3G systems faced fundamental constraints:

1. CDMA Inter-Cell Interference: Frequency reuse of 1 meant all cells interfered with each other, limiting practical capacity and cell-edge performance.

2. Complex Radio Planning: Soft handoff, power control sensitivity, and code planning created operational complexity.

3. Latency Constraints: Even optimized HSPA achieved ~50-100 ms round-trip—inadequate for emerging real-time applications.

4. Spectrum Efficiency Ceiling: CDMA approaches Shannon limits but OFDMA proved more efficient for wide channels with frequency-selective fading.

5. Voice/Data Integration: Dual-mode CS/PS cores maintained complexity; all-IP simplified architecture beckoned.

The next page explores 4G LTE, which addressed these limitations through OFDMA modulation, all-IP flat architecture, and other innovations that delivered true mobile broadband.