Limitations of Early 2G Data Services
The second generation (2G) of mobile technology, led by the Global System for Mobile Communications (GSM), brought digital voice communication, security improvements, and the Short Message Service (SMS) to wide use. It provided a reliable and relatively affordable way to communicate wirelessly across countries and continents. By the late 1990s, internet use was growing rapidly and the limitations of a voice-centred architecture for data services became increasingly apparent.
GSM was designed around voice, which is a continuous stream of information. Data services were added through CSD, GPRS, and EDGE within the existing architecture. These mechanisms provided packet and circuit data capabilities, while the mobile internet experience of the 2G era remained characterized by:
- Low Data Rates: Even with enhancements like GPRS and EDGE, data rates were measured in kilobits per second. Loading a simple webpage with images could take a long time, while streaming video or music required conditions beyond the usual 2G service profile.
- High Latency: Latency, the delay between sending a request and receiving a response, was high. This made interactive applications like online gaming or real-time communication feel slow to respond.
- Circuit-Oriented Data Architecture: The underlying GSM network was built around , which suits voice but uses resources inefficiently for bursty, intermittent internet traffic. These limits motivated a radio and core architecture with stronger support for packet data.
The Vision for Third-Generation Mobile Services
In response to these limitations, the global telecommunications community, led by the International Telecommunication Union (ITU), formulated a vision for the next generation of mobile technology. This initiative, known as IMT-2000, defined goals for higher data rates, global mobility, and a unified service framework that could support a "mobile information society".
The core objective was to create a network architecture that could handle traditional voice services together with high-speed packet data services. Several 3G standards emerged from this effort. The Universal Mobile Telecommunications System (UMTS) became a major 3G successor to GSM in the networks that adopted it.
What is UMTS? An Evolution of GSM
The Universal Mobile Telecommunications System (UMTS) is a third-generation (3G) cellular technology designed as a planned evolution of the GSM family. This approach was a key decision made by the 3GPP (3rd Generation Partnership Project), the collaboration of standards bodies that developed UMTS. It allowed mobile operators with established GSM infrastructure to upgrade to 3G in stages and reuse selected core-network functions.
The core innovation of UMTS was the redesign of the radio access technology. It reused selected GSM core-network components, such as the MSC and HLR, and introduced a new air interface with higher capacity for mobile data services. The resulting radio access network supported the data rates and service profiles targeted by IMT-2000.
UMTS's Key Improvements over 2G
UMTS reorganized the mobile network around a wider-band radio access network and service-specific capabilities. The main improvements concerned radio access, packet data, data rates, security, and Quality of Service.
A central advancement in UMTS was its adoption of a new air interface called . It used direct-sequence spreading and code separation instead of the FDMA/TDMA arrangement used in GSM.
- Wide Channels: Unlike GSM's narrow channels, WCDMA uses a much wider channel bandwidth of .
- Simultaneous Transmission: In WCDMA, users share a configured carrier at the same time and separate their transmissions with spreading codes and power control. A UTRA FDD carrier has a bandwidth of approximately .
- Code-Based Separation: The system separates simultaneous transmissions by assigning each user a spreading code. The code spreads the user's data across the configured carrier, and the receiver correlates the signal with the same code. Signals using other codes contribute interference that is managed through power control and receiver processing. UTRA also supports FDD and TDD modes, so the exact carrier arrangement depends on deployment. This arrangement supported spectral efficiency. Receiver processing could mitigate multipath effects, and soft handover let a phone connect to a new cell before releasing the old radio link.
UMTS retained a circuit-switched path for traditional voice services and introduced a native domain for data services. Packet resources could be allocated according to traffic demand, while a packet-session context could remain established between transmissions. Whether billing used time or volume depended on the operator service model, with volume-based plans becoming common for mobile data.
The initial release of the standard (Release '99) defined reference peak data rates up to for stationary or indoor use and up to for mobile use in the relevant IMT-2000 profile. Compared with approximate GPRS values around and EDGE values around , the higher rate range supported additional mobile applications. Actual values depended on the environment, UE category, radio conditions, and network configuration:
- Mobile video calling and streaming.
- Rich, graphical web browsing.
- Music downloads and streaming services.
- Location-based services with detailed maps.
UMTS extended GSM's security model with mutual authentication, integrity protection, and AKA-derived keys.
- Mutual Authentication: UMTS AKA authenticates the subscriber to the network and lets the USIM verify the serving network. Sequence numbers help the USIM and home network verify the freshness of authentication data. This addresses a GSM security limitation and strengthens resistance to rogue base-station procedures. Protection still depends on the selected access technology and the procedures supported by the network and device.
- Stronger Algorithms: UMTS used cryptographic algorithms such as KASUMI with a 128-bit algorithm key. The AKA procedures derive 128-bit ciphering and integrity keys, while individual authentication fields have their own defined sizes.
- Integrity Protection: UMTS introduced integrity protection for signalling messages. A message authentication code at the radio-resource-control layer helps the receiver detect changes to control messages exchanged by the phone and the network. The f9 algorithm produces a 32-bit MAC-I.
UMTS introduced a Quality of Service framework that classifies traffic according to application requirements. The framework defines four distinct classes. Conversational and streaming traffic are treated as real-time profiles, while interactive and background traffic follow different delay expectations:
- Conversational Class: For real-time, two-way services like voice and video calls, which are highly sensitive to delay.
- Streaming Class: For one-way real-time services like streaming video or audio, where timing is important but some delay is acceptable.
- Interactive Class: For "request-response" traffic like web browsing, where preserving the overall data integrity is more important than a perfectly fixed delay.
- Background Class: For non-time-sensitive traffic like email downloads or file transfers, which can tolerate longer delays.
The QoS classes let operators map traffic to suitable delay, ordering, and delivery requirements and allocate network resources according to the service profile. FDD and TDD profiles can apply different radio configurations.