High Speed Packet Access
High Speed Packet Access (HSPA) is a set of mobile broadband technologies developed as enhancements to the 3G Universal Mobile Telecommunications System (UMTS) to deliver higher packet data rates and improved efficiency for downlink and uplink communications. It encompasses High Speed Downlink Packet Access (HSDPA), introduced in 3GPP Release 5 in 2002, which achieves peak downlink data rates of up to 14 Mbit/s through shared channel transmission, adaptive modulation and coding (including QPSK and 16-QAM), fast hybrid automatic repeat request (HARQ), and a short 2 ms transmission time interval (TTI).[1] High Speed Uplink Packet Access (HSUPA), added in 3GPP Release 6 in 2004, complements HSDPA by supporting peak uplink rates of up to 5.76 Mbit/s using similar techniques like fast scheduling at the Node B and enhanced dedicated channel (E-DCH) operation.[2] HSPA significantly boosts spectral efficiency over legacy UMTS, enabling lower latency and greater capacity for data-intensive applications such as web browsing and file downloads on mobile devices.[3] Key architectural changes include shifting control from the radio network controller (RNC) to the base station (Node B) for faster link adaptation and scheduling, which reduces overhead and supports multiple users via time-division multiplexing on shared channels.[1] By 2010, HSPA had become a cornerstone of global 3G deployments, with widespread commercial availability enhancing mobile internet access before the transition to 4G LTE.[4] Subsequent evolutions, often referred to as HSPA+ or Evolved HSPA, were specified in later 3GPP releases (7–9), incorporating higher-order modulation (64-QAM), multiple-input multiple-output (MIMO) antennas, and dual-carrier operation to push downlink speeds to 42 Mbit/s and beyond, while maintaining backward compatibility with earlier HSPA devices.[3] These advancements improved user throughput in real-world scenarios, with average downlink rates often reaching 1–5 Mbit/s depending on network load and terminal category, and facilitated the offloading of voice services via VoIP over packet data.[5] Overall, HSPA bridged the gap between basic 3G and modern broadband, influencing over a billion mobile connections worldwide by the mid-2010s.[6]Introduction
Definition and Purpose
High Speed Packet Access (HSPA) is a collection of mobile telephony protocols defined by the 3rd Generation Partnership Project (3GPP) that enhance the Universal Mobile Telecommunications System (UMTS) for high-speed packet data transmission. It amalgamates High Speed Downlink Packet Access (HSDPA), which focuses on improving downlink performance, and High Speed Uplink Packet Access (HSUPA), which targets uplink enhancements, while HSPA+ represents the evolved version incorporating advanced features like multiple-input multiple-output (MIMO) and higher-order modulation.[7][2] The primary purpose of HSPA is to boost packet data throughput, lower latency, and enable support for multimedia services such as video streaming and web browsing on existing Wideband Code Division Multiple Access (WCDMA) infrastructure. By optimizing UMTS for packet-switched data, HSPA addresses the limitations of earlier 3G systems in handling growing demand for mobile internet and data-intensive applications without requiring a complete network overhaul.[7][8] Key benefits include peak downlink speeds of up to 14.4 Mbps via HSDPA and uplink speeds of up to 5.76 Mbps via HSUPA, alongside improved spectral efficiency through the use of shared channels that dynamically allocate resources to multiple users. These enhancements provide a cost-effective upgrade path from UMTS Release 99 WCDMA networks, primarily via software updates and minimal hardware additions to base stations.[9][2][7][8]Historical Development
High Speed Packet Access (HSPA) emerged as an enhancement to the Universal Mobile Telecommunications System (UMTS) in response to the increasing demand for mobile internet services following UMTS deployments in the early 2000s, which initially offered limited data rates that struggled to meet growing user expectations for faster packet-based communications. The technology's development was driven by the need to boost downlink and uplink capacities in 3G networks, enabling more efficient handling of web browsing, email, and emerging multimedia applications without requiring a full shift to new radio access paradigms. Standardization of HSPA began with 3GPP Release 5 in 2001-2002, which introduced High Speed Downlink Packet Access (HSDPA) as the foundational element, specifying initial features like adaptive modulation up to 16QAM and peak downlink speeds of 8-10 Mbps to improve spectral efficiency over basic UMTS. This release marked the first major packet-data evolution in UMTS, with specifications frozen in September 2002.[10][11] Building on this, 3GPP Release 6 in 2004-2005 added High Speed Uplink Packet Access (HSUPA), completing the baseline HSPA framework by enhancing uplink capabilities to up to 5.76 Mbps through techniques like enhanced dedicated channels, thus addressing the asymmetry in data traffic demands.[12] Further advancements came in 3GPP Releases 7 and 8 during 2007-2009, defining HSPA+ (also known as HSPA Evolution) with key features such as 64QAM modulation for higher-order encoding and 2x2 Multiple Input Multiple Output (MIMO) to achieve downlink peaks of 21 Mbps and 28 Mbps, respectively, while optimizing for lower latency and greater capacity. These releases solidified HSPA's role as a bridge to higher-speed mobile broadband before the advent of Long Term Evolution (LTE).[13] The first commercial HSDPA launches occurred in 2006, led by Hutchison 3G operators in Europe and Asia; for instance, 3 Italia activated services in February 2006, followed by 3 UK in May and 3 Hong Kong in the third quarter of 2006, marking the initial widespread deployment of HSPA technology. By 2010, HSPA had seen rapid global adoption, with 358 networks operational in 142 countries as of September 2010, reflecting its peak expansion in the early 2010s prior to LTE's dominance.[14]Technical Foundations
UMTS Integration and Prerequisites
High Speed Packet Access (HSPA) builds upon the Universal Mobile Telecommunications System (UMTS), which serves as the foundational third-generation (3G) mobile communication standard developed under the International Mobile Telecommunications-2000 (IMT-2000) framework. UMTS employs Wideband Code Division Multiple Access (W-CDMA) as its air interface technology, utilizing 5 MHz channel carriers to enable higher data rates and improved spectral efficiency compared to second-generation systems like GSM.[15] Typically deployed in the 2100 MHz frequency band for initial global rollouts, UMTS supports both circuit-switched and packet-switched services, laying the groundwork for enhanced packet data capabilities in HSPA. This structure allows HSPA to extend UMTS without requiring a complete overhaul of the existing infrastructure. Key prerequisites for understanding HSPA integration include the core elements of UMTS network architecture and channel configurations. The UMTS Terrestrial Radio Access Network (UTRAN) comprises Node Bs, which function as base stations handling radio transmission and reception with user equipment (UE), and Radio Network Controllers (RNCs), which manage resource allocation, handover decisions, and overall radio resource control across multiple Node Bs.[16] In UMTS, channels are categorized as dedicated or shared: dedicated channels, such as the Dedicated Transport Channel (DCH), allocate resources exclusively to a single UE for continuous services like voice, while shared channels, like the Downlink Shared Channel (DSCH), allow multiple UEs to access the same resource dynamically for bursty data traffic, promoting efficiency in resource utilization.[17] These concepts are essential, as HSPA leverages and modifies the shared channel paradigm to achieve higher speeds. HSPA integrates seamlessly with the UMTS air interface by reusing its fundamental W-CDMA framework, including spreading, modulation, and multiplexing techniques, while introducing specialized high-speed shared channels to boost packet data performance. Specifically, the High-Speed Downlink Shared Channel (HS-DSCH) enhances downlink packet transmission, and the Enhanced Dedicated Channel (E-DCH) improves uplink capabilities, both operating within the existing UMTS physical layer structure.[18] This approach ensures that HSPA enhancements are applied as an overlay on UMTS carriers, maintaining compatibility with the core network elements like the RNC and circuit-switched domain. Regarding frequency bands, HSPA primarily supports Frequency Division Duplex (FDD) mode in globally allocated UMTS spectrum, including 850 MHz, 900 MHz, 1900 MHz, and 2100 MHz, enabling widespread deployment across regions without spectrum reallocation. Backward compatibility is a cornerstone of HSPA design, allowing HSPA-capable devices to revert to basic UMTS operations for services not optimized by packet enhancements, particularly voice and other circuit-switched applications. When packet data channels are unavailable or unsuitable, UEs fall back to UMTS dedicated channels for reliable circuit-switched voice transmission, ensuring uninterrupted service in mixed-network environments.[19] This fallback mechanism supports coexistence of legacy UMTS terminals and newer HSPA devices on the same carrier, facilitating gradual network upgrades.[20]Key Enabling Technologies
High Speed Packet Access (HSPA) relies on several core innovations that enhance packet data efficiency over UMTS networks, enabling higher throughputs while maintaining compatibility with existing infrastructure. These technologies include adaptive modulation and coding, hybrid automatic repeat request, fast scheduling at the Node B, multiple input multiple output in later evolutions, and refined power control mechanisms. Together, they address channel variability, error recovery, resource allocation, and interference, forming the foundation for both downlink and uplink enhancements.[21] Adaptive Modulation and Coding (AMC) dynamically adjusts the modulation scheme and coding rate based on instantaneous channel quality feedback from the user equipment, selecting between QPSK for robust transmission in poor conditions and 16QAM for higher spectral efficiency in favorable ones. This link adaptation maximizes throughput by matching the transmission parameters to the radio environment, reducing wasted capacity on unreliable links. In HSPA, AMC operates on a per-transmission time interval (TTI) basis, with the Node B using channel quality indicators (CQI) reported by the UE to select the optimal modulation and coding scheme (MCS).[22][21] Hybrid Automatic Repeat Request (HARQ) integrates forward error correction (FEC) with automatic repeat request (ARQ) protocols, employing incremental redundancy to retransmit only additional parity bits for failed packets rather than entire blocks. This approach enables faster error recovery and higher reliability, supporting up to eight parallel HARQ processes in the downlink to pipeline retransmissions without stalling data flow. HARQ operates at the physical layer within the Node B, combining soft-decoded bits from initial and subsequent transmissions to improve decoding success rates.[23][21] Fast scheduling at the Node B allocates shared channel resources every 2 ms TTI to users experiencing the best channel conditions, prioritizing those with high carrier-to-interference ratios to optimize overall cell throughput. This Node B-centric approach, unlike the slower radio network controller-based scheduling in earlier UMTS, reduces latency and exploits multi-user diversity by serving multiple users opportunistically within the available code and power resources.[21] In HSPA+ (3GPP Release 7 and beyond), Multiple Input Multiple Output (MIMO) introduces spatial multiplexing using two transmit and two receive antennas, effectively doubling the peak data capacity by transmitting independent data streams over the same frequency band to achieve up to 28 Mbps using 16QAM. Separately, higher-order modulation like 64QAM supports up to 21.6 Mbps in single-antenna configurations. These features improve spectral efficiency in multi-path environments.[13] The theoretical peak throughput in HSPA can be approximated using the formula: \text{Peak Throughput} = (\text{bits per symbol}) \times (\text{symbols per TTI}) \times (\text{TTIs per second}) For HSDPA with 16QAM (4 bits/symbol), 7200 symbols per 2 ms TTI, and 500 TTIs per second, this yields approximately 14.4 Mbps, representing the maximum under ideal conditions with full code utilization.[24] Power control and interference management in HSPA build on UMTS inner-loop mechanisms, where the Node B adjusts UE transmit power in 1500 Hz steps (1.5 dB) to maintain a target signal-to-interference ratio, preventing dominance by strong users and mitigating intra-cell interference. Enhancements for packet data include fractional dedicated physical control channel (F-DPCH) to reduce pilot overhead and outer-loop adjustments for varying quality of service targets, ensuring efficient resource use across shared channels.[25]Downlink Enhancements
HSDPA Core Mechanisms
High Speed Downlink Packet Access (HSDPA) employs the High-Speed Downlink Shared Channel (HS-DSCH) as its primary transport channel for user data transmission, utilizing a fixed spreading factor of 16 to support up to 16 channelization codes, with a maximum of 15 codes multiplexed for allocation to a single user equipment (UE) within a transmission time interval (TTI).[26] This code multiplexing enables efficient sharing of downlink resources among multiple users while maintaining compatibility with the underlying UMTS code-division multiple access framework. At the medium access control (MAC) layer, the MAC-hs protocol entity, located in the Node B, oversees key functions including packet scheduling based on channel conditions, Hybrid Automatic Repeat reQuest (HARQ) for error correction and retransmissions, and dynamic rate adaptation to optimize throughput.[27] These mechanisms allow the MAC-hs to prioritize transmissions for UEs with favorable radio conditions, leveraging brief references to HARQ and adaptive modulation and coding for rapid adjustments without deeper protocol details. The transport block size on the HS-DSCH varies dynamically up to a maximum of 27,952 bits per 2 ms TTI (for Category 10), which may be segmented into multiple code blocks of up to 5,114 bits each, determined through outer-loop link adaptation that targets a specific block error rate by adjusting based on feedback from prior transmissions.[28][8] Supporting control channels facilitate the HSDPA data flow: the High-Speed Shared Control Channel (HS-SCCH) carries downlink scheduling information, such as modulation scheme, transport block size, and code allocation, enabling the UE to decode the subsequent HS-DSCH transmission.[1] Conversely, the High-Speed Dedicated Physical Control Channel (HS-DPCCH) in the uplink conveys the UE's acknowledgment (ACK) or negative acknowledgment (NACK) for received blocks, along with Channel Quality Indicator (CQI) reports to inform the Node B of current downlink conditions.[1] The end-to-end process begins with the UE transmitting CQI feedback every TTI via the HS-DPCCH, allowing the Node B's MAC-hs scheduler to select the appropriate modulation, coding, and resource allocation for the next HS-DSCH transmission.[29] Upon reception, the UE processes the data and responds with ACK/NACK; successful ACKs confirm delivery, while NACKs trigger HARQ-based retransmissions from the Node B, often with incremental redundancy to enhance reliability. This closed-loop operation ensures efficient resource use and error recovery at the physical layer.[1] By shifting scheduling and HARQ processing to the Node B with a shortened 2 ms TTI, HSDPA achieves a significant latency reduction, lowering round-trip times from 100-150 ms in Release 99 UMTS to approximately 50 ms through faster feedback and adaptation cycles.[8]HSDPA User Equipment Categories
HSDPA User Equipment (UE) categories, defined by the 3GPP starting in Release 5, classify devices based on their support for key downlink parameters such as the maximum number of HS-DSCH codes, modulation schemes, and peak data rates. These categories enable networks to allocate resources appropriately, with higher categories offering enhanced capabilities for faster packet access while maintaining compatibility with earlier infrastructure. The specifications are detailed in 3GPP TS 25.306, which outlines the radio access capabilities for each category. Categories 1 through 6, introduced in Release 5, provide foundational support for downlink speeds with QPSK and 16-QAM modulation. Categories 1-4 utilize 5 HS-DSCH codes for peak rates up to 1.8 Mbps, while categories 5-6 extend to 10 codes for peaks up to 3.6 Mbps. These were suited for initial deployments targeting voice and low-data applications.[30] Categories 7-8 (Release 5) and 9-10 (Release 6) advance capabilities with up to 15 HS-DSCH codes and QPSK/16-QAM modulation for peak rates up to 14.4 Mbps. Category 8 uses 10 codes to achieve 7.2 Mbps, while category 10 uses 15 codes for 14.4 Mbps, enabling better spectral efficiency in moderate network loads. These mid-range categories balanced cost and performance for emerging data-intensive uses like web browsing. From Release 7 onward, categories 13-14 incorporate 64-QAM modulation to reach peak rates of 21.1 Mbps on single carriers, with further enhancements in later releases adding MIMO and multi-carrier operation. Category 14 supports 64-QAM with 15 codes for 21.1 Mbps, marking a significant evolution for high-definition video and file transfers. Key parameters across categories include the maximum HS-DSCH codes (ranging from 5 to 15 initially, expanding in multi-carrier setups), supported modulations (QPSK to 64-QAM), and MIMO (introduced in categories 15+ for spatial multiplexing). For example, category 24 achieves 42 Mbps through dual-carrier HSDPA and 2x2 MIMO, combining two 5 MHz carriers with 64-QAM. These advanced features are specified in 3GPP TS 25.101 for radio transmission requirements.[31] Backward compatibility is inherent in the design, allowing higher-category UEs to negotiate down to the network's supported capabilities during connection setup, ensuring seamless operation across diverse deployments. Representative examples include early USB modems often implementing category 6 for portable broadband access, and modern smartphones adopting category 10 or higher to leverage 16-QAM for improved download speeds. The following table summarizes select HSDPA UE categories, focusing on peak downlink rates, key parameters, and release introductions (theoretical peaks under ideal conditions; actual rates vary with channel quality):| Category | Release | Max HS-DSCH Codes | Modulation Support | Peak Rate (Mbps) | Notable Features |
|---|---|---|---|---|---|
| 1–4 | 5 | 5 | QPSK, 16-QAM | Up to 1.8 | Basic packet access |
| 5–6 | 5 | 10 | QPSK, 16-QAM | Up to 3.6 | Extended codes for higher throughput |
| 8 | 5 | 10 | QPSK, 16-QAM | 7.2 | Mid-range with 16-QAM |
| 9–10 | 6 | 15 | QPSK, 16-QAM | 10.1–14.4 | Maximum single-carrier pre-64QAM |
| 13–14 | 7 | 15 | QPSK, 16-QAM, 64-QAM | 17.6–21.1 | 64-QAM for single-carrier boost |
| 24 | 8+ | 32 (dual-carrier) | 64-QAM, MIMO | 42 | Dual-carrier and 2x2 MIMO |
HSDPA Performance Characteristics
High Speed Downlink Packet Access (HSDPA) achieves a theoretical peak downlink data rate of 14.4 Mbps in user equipment Category 10, utilizing 16 quadrature amplitude modulation (16QAM), 15 orthogonal codes, and higher-order channel coding.[32] This peak performance assumes ideal channel conditions, including high signal-to-noise ratio (SNR) and minimal interference, as defined in 3GPP Release 6 specifications.[33] In real-world deployments, however, typical downlink throughputs range from 1 to 7 Mbps, influenced by factors such as multipath fading, inter-cell interference, network loading, and mobility.[34] Adaptive modulation and coding (AMC) plays a critical role here, dynamically adjusting modulation schemes and coding rates based on instantaneous SNR to optimize throughput; lower SNR at cell edges or in high-mobility scenarios often reverts to quadrature phase-shift keying (QPSK), reducing rates to around 384 kbps.[1] HSDPA significantly enhances spectral efficiency compared to baseline UMTS, reaching up to 0.69 bits per second per hertz (bps/Hz) in downlink cell throughput, versus UMTS's approximately 0.2 bps/Hz, primarily through fast scheduling and shared channel resources that allocate power efficiently among users.[35] Early field trials in 2006, such as those conducted with Category 6 devices supporting 3.6 Mbps peaks, demonstrated average throughputs exceeding 1 Mbps in median conditions, with lab validations confirming up to 3.6 Mbps under controlled loads.[36] Power efficiency is improved via the shared High-Speed Downlink Shared Channel (HS-DSCH), which reduces transmit power per bit by concentrating resources on active users rather than dedicating channels, leading to lower overall base station energy consumption per delivered bit compared to Release 99 UMTS.[37]Uplink Enhancements
HSUPA Core Mechanisms
High Speed Uplink Packet Access (HSUPA) introduces the Enhanced Dedicated Channel (E-DCH) as its primary uplink transport channel, which operates as a hybrid of dedicated and shared channel access to enable efficient packet data transmission from the user equipment (UE) to the network.[38] The E-DCH utilizes separate Code Division Test Resource Channel (CCTrCH) structures from dedicated channels, supporting one transport block per transmission time interval (TTI) and employing turbo coding with a 24-bit cyclic redundancy check (CRC) for error detection.[38] The E-DCH employs two key physical channels: the E-DCH Dedicated Physical Data Channel (E-DPDCH) for carrying user data and the E-DCH Dedicated Physical Control Channel (E-DPCCH) for transmitting control signaling, including the E-DCH Transport Format Combination Indicator (E-TFCI), retransmission sequence number (RSN), and a "Happy Bit" indicating UE buffer status satisfaction.[39] This separation allows precise control over data transmission while minimizing overhead. Scheduling for the E-DCH is performed at the Node B to manage uplink resources and interference, using absolute grants sent via the E-DCH Absolute Grant Channel (E-AGCH) to set the maximum allowed E-DPDCH-to-DPCCH power ratio for rate control, and relative grants transmitted on the E-DCH Relative Grant Channel (E-RGCH) to incrementally adjust this ratio (e.g., "UP," "DOWN," or "HOLD") for interference coordination across serving and non-serving radio links.[39][40] The Node B scheduler considers UE-provided quality of service (QoS) information from the serving radio network controller (SRNC) alongside scheduling requests to allocate resources.[40] HSUPA supports configurable TTI lengths of 2 ms or 10 ms, with the shorter 2 ms option enabling lower latency for delay-sensitive applications through faster hybrid automatic repeat request (HARQ) round-trip times (approximately 16 ms versus 40 ms for 10 ms TTI), while the longer 10 ms TTI improves coverage in challenging radio conditions.[39] This flexibility is complemented by 4 HARQ processes for 10 ms TTI or 8 for 2 ms TTI, allowing continuous transmission during retransmissions.[39] At the medium access control (MAC) layer, the MAC-es and MAC-e entities in the UE manage E-DCH-specific operations, including segmentation of MAC-d protocol data units (PDUs) from logical channels into MAC-es PDUs, multiplexing of these MAC-es PDUs into MAC-e PDUs per TTI, and HARQ handling for reliable delivery.[39] The E-TFC selection process at the UE determines the appropriate transport format based on available grants, buffer occupancy, and power constraints before transmission.[39] The operational process begins with the UE sending scheduling information—such as total E-DCH buffer status (TEBS), highest priority logical channel ID (HLID), highest priority buffer status (HLBS), and non-scheduled transmission power—via the E-DPCCH or dedicated control channel to request resources.[39] The Node B scheduler evaluates these requests and issues absolute or relative grants to allocate uplink capacity, after which the UE transmits data on the E-DPDCH accordingly.[40] Fast retransmissions occur via the configured HARQ processes if acknowledgments (ACK/NACK) received on the E-DCH HARQ Acknowledgment Indicator Channel (E-HICH) indicate errors, ensuring efficient error recovery without upper-layer intervention.[39] In 3GPP Release 7, uplink enhancements include support for 16QAM modulation on the E-DPDCH, enabling higher spectral efficiency and peak data rates, alongside the 2 ms TTI to reduce round-trip latency to approximately 60 ms for typical small-packet transmissions.[39]HSUPA User Equipment Categories
High Speed Uplink Packet Access (HSUPA) User Equipment (UE) categories define the uplink capabilities of devices supporting the Enhanced Dedicated Channel (E-DCH), as specified in 3GPP technical standards. These categories vary in maximum bit rates, number of channelization codes, supported Transmission Time Intervals (TTI), modulation schemes, and other parameters that determine the peak uplink throughput and compatibility with network resources. Early categories, introduced in 3GPP Release 6, focused on basic enhancements over UMTS uplink, while later releases added advanced features like higher-order modulation and multi-carrier support to achieve greater speeds.[41] The primary parameters for HSUPA UE categories include the maximum E-DCH transport block size (which directly influences bit rate), the number of E-DCH channelization codes transmitted per TTI, the minimum spreading factor, TTI length (10 ms or 2 ms), supported modulation (starting with QPSK and evolving to 16QAM or 64QAM), and the ability to support simultaneous voice services via circuit-switched channels alongside packet data. For instance, categories supporting a 2 ms TTI enable lower latency and higher peak rates compared to 10 ms TTI, but require more robust UE hardware. Additionally, some categories allow concurrent operation with Dedicated Channels (DCH) for voice, ensuring compatibility with legacy UMTS services.[41] HSUPA UEs are required to be dual-mode, meaning they must also support High Speed Downlink Packet Access (HSDPA) to enable full HSPA functionality, as the standards integrate uplink and downlink enhancements within the same radio access capabilities framework. This ensures seamless operation in networks deploying both technologies. Early implementations, such as Category 1 to 3 UEs, achieved peak uplink rates up to approximately 1.4 Mbps using QPSK modulation, 1 to 2 codes (spreading factors of 4), and primarily 10 ms TTI, with limited simultaneous voice support in some configurations; these were common in initial USB dongles and modems for basic mobile broadband.[41] Subsequent categories 5 and 6, also from Release 6, improved upon this with peak rates up to 5.76 Mbps, employing up to 4 codes (2 at SF=2 and 2 at SF=4), QPSK modulation, and 2 ms TTI support for enhanced efficiency, while maintaining compatibility for simultaneous voice and data. In Release 7 and beyond, higher categories like 7 through 9 introduced 16QAM modulation alongside QPSK, enabling peak rates up to 11.5 Mbps in single-carrier mode with 4 codes and 2 ms TTI; Category 9, for example, supports advanced coding and can reach up to 22 Mbps in dual-cell configurations under Release 9, prioritizing high-throughput applications in modern smartphones and data cards. These evolutions prioritize backward compatibility while scaling uplink performance for diverse use cases.[41] The following table summarizes key HSUPA UE categories, focusing on representative examples across releases (calculated peak bit rates based on maximum transport block sizes and TTI):| Category | Release | Peak Uplink Bit Rate (Mbps) | Max E-DCH Codes | Modulation | TTI (ms) | Simultaneous Voice Support |
|---|---|---|---|---|---|---|
| 1 | 6 | 0.73 | 1 (SF=4) | QPSK | 10 | Yes (limited) |
| 2 | 6 | 1.45 (10 ms) / 1.4 (2 ms) | 2 (SF=4) | QPSK | 10 & 2 | Yes |
| 3 | 6 | 1.45 | 2 (SF=4) | QPSK | 10 | Yes |
| 5 | 6 | 2.0 | 2 (SF=2) | QPSK | 10 | Yes |
| 6 | 6 | 5.76 (2 ms) | 4 (2xSF=2, 2xSF=4) | QPSK | 10 & 2 | Yes |
| 7 | 7 | 11.5 (2 ms) | 4 (SF=2) | QPSK, 16QAM | 10 & 2 | Yes |
| 9 | 7+ | 11.5 (single-cell) / 22 (dual-cell) | 4 (SF=2) | QPSK, 16QAM | 2 | Yes |