Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from F-J


uplink downlink rate code

Athanassios C. Iossifides
COSMOTE S.A., Greece

Spiros Louvros
COSMOTE S.A., Greece

Stavros A. Kotsopoulos
University of Patras, Greece


Global rendering of personalized multimedia services is the key issue determining the evolution of next-generation mobile networks. The determinant factor of mobile multimedia communications feasibility is the air-interface technology. The Universal Mobile Telecommunications System (UMTS) evolution, based on wideband code-division multiple access (WCDMA), constitutes a major step to the target of truly ubiquitous computing: computing anywhere, anytime, guaranteeing mobility and transparency. However, certain steps are still required in order to achieve the desired data rates, capacity, and quality of service (QoS) of different traffic classes inherent in multimedia services.

A view of data-rate trends of applied and future mobile communications technologies is shown in Figure 1. UMTS, being in its premature application stage, is currently providing rates up to 64/384 Kbps (uplink [UL]/downlink [DL]). It was initially designed to provide rates up to 2 Mbps under ideal conditions, which seems not enough from a competitiveness point of view compared to WLANs (wireless local-area networks) that aim to easily reach 2-to 10-Mbps data rates with the possibility of reaching 100 Mbps (Simoens, Pellati, Gosteau, Gosse, & Ware, 2003). Hardware, software, installation, and operational costs of 3G (3rd Generation) systems could be proven unjustified and unprofitable if they cannot cope with at least a certain share of data rates over 2 Mbps. This article focuses on the characteristics, application, and future enhancements (planned in 3GPP Release 5 and 6 or under research) of WCDMA-FDD (frequency-division duplex) toward high-quality multimedia services.


CDMA, in contrast to FDMA (Frequency Division Multiple Access) and TDMA (Time Division Multiple Access), poses no restrictions to the time interval and frequency band to be used for the transmission of different users. All users can transmit simultaneously while occupying the whole available bandwidth (Figure 2). They are separated by uniquely (per user) assigned codes with proper low cross-interference properties. Thus, while interuser interference is strictly avoided in TDMA and FDMA systems by assigning different portions of time (time slots [TSs]) or bandwidth to different users, respectively, interuser interference, referred to as multiple-access interference (MAI), is inherent in CDMA techniques and is the limiting capacity factor (interference-limited systems).

Although CDMA has been known for several decades, only in the last two decades has interest peaked regarding its use for mobile communications because of its enhanced performance compared to standard TDMA and FDMA techniques. Greater capacity, exploitation of multipath fading through RAKE combining, soft handover, and soft capacity are some of CDMA’s advantages (Viterbi, 1995). The first commercial CDMA mobile application was IS-95 (1993). The real boost of CDMA applications, though, was the adoption of the WCDMA air interface for UMTS.

CDMA is applied using spread-spectrum techniques, such as frequency hopping (FH), direct sequence (DS), or hybrid methods. The DS technique, which is used in UMTS, is applied by multiplying the information symbols with faster pseudorandom codes of low cross-correlation between each other, which spreads the information bandwidth (Figure 3). The number of code pulses (chips) used for spreading an information symbol is called the spreading factor (SF). The higher the SF, the greater the tolerance to MAI is. A simplified block diagram of a CDMA transmitter and receiver is given in Figure 4. The receiver despreads the signal with the specific user’s unique code followed by an integrator or digital summing device. Coexistent users’ signals act as additive wideband noise (MAI).

With properly selected codes (of low autocorrelation), multipath propagation can turn into diversity gain for CDMA systems as soon as multiple paths’ delays are spaced more than the chip duration (these paths are called resolved). In such a case, a RAKE receiver is employed (Figure 5), which performs a full reception procedure for each one of the resolved paths and properly combines the received signal replicas. In any case, discrimination between CDMA users is feasible with conventional receivers (no multiuser receivers) only when an advanced power-control method is engaged. Otherwise the near-far effect will destroy multiple-access capability.

There is no universally accepted definition for what is called WCDMA. From a theoretical point of view, a CDMA system is defined as wideband when the overall spread bandwidth exceeds the coherence bandwidth of the channel (Milstein, 2000). In such a case, the channel appears to be frequency selective, and multipath resolvability is possible. Compared to narrowband CDMA, beyond multipath exploitation, WCDMA presents enhanced performance through certain advantages, such as a decrease of the required transmitted power to achieve a given performance, greater tolerance to power-control errors, fading-effects reduction, the capability to transmit higher data rates and multimedia traffic, and so forth.


This section summarizes the basic concepts and procedures of applied WCDMA-FDD systems (based on 3GPP [3rd Generation Partnership Project] Rel. 99 or at most Rel. 4).

Information Organization

Source information arrives in transmission time intervals (TTIs) of 10, 20, 40, or 80 ms. Information bits are organized in blocks and CRC (Cyclic Redundancy Check) attachment, forward error-correction (FEC) coding, rate matching, interleaving, and information multiplexing are applied (Holma & Toskala, 2000). FEC can be convolutional of rate 1/2, 1/3, or turbo of rate 1/3, depending on the information type. The produced channel symbols are of rate 7.5*2 m ( m = 0 to 7) Ksps. Examples of coding and multiplexing are given in 3GPP TR (Technical Report) 25.944.

Multiple-Access Methodology

Multiple access is realized through channelization and scrambling. Channel symbols are spread by the channelization code (orthogonal Hadamard codes) and then chip-by-chip multiplication with the scrambling code takes place (long, partial gold codes of length 38,400 or short S(2) codes of length 256 for future uplink multiuser reception). The chip rate of both channelization and scrambling codes is constant at 3.84 Mchip/s.

In the uplink, each user is assigned a unique (among 2 24 available) scrambling code. Channelization codes are used for separating data and control streams between each other and may have lengths and SFs equal to 2 k ( k = 2 to 8) chips. Data-rate variability is achieved by alternating the length of the channelization code (SF) that spreads information symbols. The greater the SF, the lower the information rate is, and vice versa. Parallel usage of more than one channelization code for high uplink data rates (multicode operation) is allowed only when SF equals 4, and this has not been commercially applied yet.

Downlink separation is twofold. Cells are distinguished between each other by using different primary scrambling codes (512 available). Each intracell user is assigned uniquely an orthogonal channelization code (SF = 2 k , k = 2 to 9). The same channelization-code set is used in every cell (Figure 6). In order to achieve various information rates (by different SFs) while preserving intracell orthogonality, the channelization codes form an orthogonal variable spreading factor (OVSF) code tree. While codes of equal length are always orthogonal, different length codes are orthogonal under the restriction that the longer code is not a child of the shorter one. Such cases are displayed in Figure 7. Additional scrambling codes (secondary) can be used in the cell mainly for enhancing capacity by reusing the channelization codes. Table 1 summarizes commercially available radio-access bearers (RABs) for circuit-switched (CS) and packet-switched (PS) services.


Data transmission is organized in frames of 10 ms (38,400 chips) consisting of 15 TSs (2,560 chips). Each TS contains information data (DPDCH — Dedicated Physical Data Channel) and physical-layer signaling (DPCCH — Dedicated Physical Control Channel;Figure 8). DPDCH and DPCCH are quadrature multiplexed before being scrambled in the uplink and time multiplexed in the downlink. Modulation at the chip level is quadrature phase shift keying (QPSK) in both uplink and downlink. Demodulation is coherent.

Soft Handover

Soft handover is the situation when a mobile station communicates with more than one cell in parallel, receiving and transmitting identical information from and to the cells (Figure 9). The cells serving the MS consist of the active set and may belong to the same (softer handover) or other BSs. A combination of cells’ signals takes part in the MS RAKE receiver in the downlink, and in the BS RAKE (in the softer case) or in the RNC (radio network controller; signal selection through CRC error counting) in the uplink. Gains in the order of 2 dB (in the signal-to-interference ratio [SIR]) have been reported with soft handover, resulting in enhanced-quality reception. The drawback is the consumption of the limited downlink orthogonal-code resources of the active-set cells.

Power Control

Fast power control (1,500 Hz) is very important for overcoming the near-far problem, reducing power consumption for acceptable communication, and eliminating fading by a significant amount for relatively low-speed moving MSs. Fast power control is based on achieving and preserving a target SIR value (set with respect to information type). An outer power control is used in the uplink for adjusting the SIR target to catch the MSs speed and environmental changes. Power-control-balancing techniques are employed in the downlink to prevent large power drifts between cells during soft handover.

Capacity, Coverage, and Information-Rate Considerations

The capacity and coverage of the system are dynamically adjusted according to the specific conditions and load. Rather involved RRM (radio resource management), admission, and congestion-control strategies have been developed to guarantee acceptable quality of service. In any case, there are some limiting factors that need to be addressed regarding the capacity and information-rate capability of the system.

Speaking of the uplink, 64 Kbps (circuit or packet) is the achievable standard information rate commercially available. Total cell throughputs in the order of 1.5 Mbps have been predicted for microcells (Holma & Toskala, 2000). Additionally, admission-control parameterizations normally assume a reception of 120 to 180 ASEs for uplink cells (air-interface speech equivalent; 1.6 ASEs for voice, 11.1 for 64 Kbps CS, 8.3 for 64 Kbps PS; Ericsson, 2003), that is, about 100 voice users or about 15 users of 64 Kbps at most. Noting that urban 2G (2nd Generation) microcells support more than 80 Erlangs of voice traffic during busy hours, 3G voice capacity seems to be adequate. However, the capacity of higher rate services still seems well below the acceptable limit for mass usage. A rate of 384 Kbps is possible when SF equals 4. Low SF, however, means low tolerance to interference and the need for a power increase. The power capability of MSs may not be enough for achieving the desired SIR when SF equals 4, especially when they are far positioned from the Node-B. Thus, the usage radius of high information rates in a fully developed network would be in the order of decades of meters and for a small number of users (because of the large amount of MAI they produce for coexistent users).

While coverage is uplink limited, capacity is clearly downlink limited (Holma & Toskala). The capacity limitation of the downlink (which should normally support higher rate services) is threefold: the BS power-transmission limitation, limited downlink orthogonality, and the cost limitation of complex MS receivers. The initiation of each new user in the system presupposes enough BS power and OVSF tree-branch availability to support the requested data. Besides this, the initiation of new users poses additional interference to other cell MS receivers. Moreover, the downlink is more sensitive to environmental differences. Although multipath may enhance performance when MSs engage adequate RAKE branches, it also leads to intracell orthogonality loss. Downlink throughputs of the order of 1.5 Mbps have been predicted (Holma & Toskala).



Transmit Diversity and MIMO Systems

Although not yet commercially applied, transmit diversity methods have been early specified for performance enhancement (space-time transmit diversity [STTD]; 3GPP TS 25.211). Each cell engages two transmit antennas and a proper coding procedure consisting of transmitting identical information in a different order and format (space-time block coding), resulting in extra diversity reception at the receiver. The system may operate in either an open-loop or a closed-loop format, where feedback information (FBI bits) is used to adjust the transmission gains of the antennas. It has been demonstrated (Bjerke, Zvonar, & Proakis, 2004; Vanganuru & Annamalai, 2003) that gains of more than 5 dB (in SNR — Signal to Noise Ratio) can be achieved with a single receiving antenna for open-loop schemes when compared to no transmission diversity. Closed-loop schemes provide an extra 3 dB gain, while engaging a second receiving antenna at the MS enhances performance by 3 to 4 dB more. Transmit diversity is a special case of the MIMO (multiple-input, multiple-output) concept that has gained great interest in the last decade (Molisch & Win, 2004). MIMO systems may be used for diversity enhancement or spatial-information multiplexing for information-rate increase. However, the implementation of multiantenna systems, especially for the MS, is still too costly. MIMO systems will play a significant role in WCDMA enhancement, but their commercial use is still far.

Advanced Receivers

The evolution of MS receivers will yield capacity enhancement of the system. Several strategies have been proposed. The key concept is to minimize orthogonality loss that arises from multipath propagation. The proposed methods employ MMSE (minimum mean squared error) receivers for chip-level equalization (Hooli, Latva-aho, & Juntti, 1999) or generalized RAKE receivers (Bottomley, Ottosson, & Wang, 2000) with tap gains derived from a maximum-likelihood criterion. These schemes produce gains in the order of 2 to 3 dB (uncoded performance) over the standard RAKE structure. Additionally, multipath interference cancellers (MPICs) have been considered (Kawamura, Higuchi, Kishiyama, & Sawahashi, 2002) with comparable performance (slightly worse) and lower complexity that reaches (in high SIRs) flat fading performance. Maximum-likelihood sequence estimation (MLSE) on the chip level is another method that guarantees even better performance at a higher complexity (Tang, Milstein, & Siegel, 2003). In any case, advanced receivers’ adoption for commercial use will pose a complexity and performance equilibrium as a selection point for manufacturers and end users (with respect to cost).

HSDPA and Link Adaptation Methods

High-speed downlink packet access (HSDPA; Honkasalo et al., 2002) is a key enhancement over the 2-Mbps packet data of WCDMA. Starting with 3GPP TR 25.848 (V4.0.0), a certain number of changes where finally adopted in Rel. 5 (3GPP TS 25.308 V.5.5.0), a brief description of which is given below. A new downlink data-traffic physical channel was introduced (HS-PDSCH — High Speed-Physical Downlink Shared Channel) with a frame of 3 TSs (2 ms), called a subframe, and a corresponding TTI, which allows faster changes of transmitted formats. SF is always set to 16. A specific part of the code tree is assigned to HSPDA (5 to 12 codes). Each HS-PDSCH is associated with a DPDCH channel, and a single information transmission to an MS is allowed in each subframe.

Adaptive modulation and coding (AMC) engaging higher order modulation schemes that span the region from QPSK to 64 QAM (quadrature amplitude modulation) regarding modulation, and 1/4 to 3/4 turbo coding, were proposed and evaluated (Nakamura, Awad, & Vagdama, 2002). The idea is to adapt the modulation-coding scheme (MCS) to the changing channel conditions using reverse-link channel-quality indicators (CQIs) in order to achieve higher throughputs. In Rel. 5 and 6 (3GPP TS 25.308 V.5.5.0, V.6.1.0) QPSK and 16 QAM were finally specified for usage, reaching a bit rate of 720 Kbps. Multicode transmission further improves information rates (Kwan, Chong, & Rinne, 2002), exceeding 2-Mbps throughputs per cell for low-speed terminals with five parallel codes.

Hybrid ARQ (autorepeat request) is engaged for enhancing performance by retransmissions (according to acknowledgement messages by the MS) when packets are received in error. Incremental redundancy (more parity bits at retransmissions) or chase-combining techniques (identical retransmission) may be used.

Fast cell selection is specified, where the users are served by the best cell of the active set at each instance, decreasing, in this way, interference, and increasing capacity.

Fast transmission scheduling is also considered, including the sequential order of serving MSs (round-robin), max C/I (Carrier to Interference) selection (where the best user is served in each TTI), or proportionally fair serving that is a trade-off between throughput maximization and fairness.

The above techniques, in conjunction with MIMO methods, promise rates that reach about 7 Mbps with advanced receiver structures (Kawamura et al., 2002). Some further enhancements under consideration (3GPP TR 25.899 V.6.0.0) include multiple simultaneous transmissions to the MS in the duration of a subframe. In this way, scheduled retransmission can be multiplexed with new transmissions to provide higher throughput. OVSF code sets can be reused in a partial (only for HSDPA codes) or full manner by using a secondary scrambling code in conjunction with two transmit antennas where each one is scheduled to transmit to specific users according to interference experienced in the downlink. Fast, adaptive emphasis for users in a soft handover with closed-loop STTD will exist, where the antenna gains are set according to the existence or nonexistence of downlink HSDPA information. Fractional, dedicated physical channels where the associated dedicated physical channels of different users can be multiplexed together in a single downlink code in order to reduce downlink code-set consumption will be implemented.


The main approaches for uplink enhancement, in terms of performance and information rate, are multiuser detectors or interference cancellers, and adaptive multirate techniques.

Increasing the complexity in BS receivers will be inevitable for increasing uplink capacity. Being out of this discussion because of its great complexity, optimal multiuser detection (MUD; Verdu, 1998) gave rise to blind, adaptive MUD approaches that avoid perfect knowledge of all user codes and channel statistics either with training sequences or without. The second approach is based on step-by-step orthogonalization (MMSE) of the desired signal to the interference by adding a proper varying vector. Newer techniques can cope with multipath interference as well as with proper precombining and windowing methods, aiming to orthogonalize the total interference signal received from all RAKE branches (Mucchi, Morosi, Del Re, & Fantacci, 2004). This method (which can be used either in the UL or DL) achieves great performance, approaching single-user behavior with near-far resistance. Other interesting methods are based on interference cancellation in conjunction with beam-forming techniques and space-time combining (Mottier & Brunel, 2003). The idea is to reproduce interference iteratively and cancel it from the desired signal. Near-single-user performance is achieved.

Multicoding has also been proposed for uplink communication. Multicode multidimensional and multicode-OVSF schemes (Kim, 2004) with proper receivers have been evaluated for reliable, high uplink transmission rates. Adaptive schemes with rate adaptation (multiple SFs) have also been analyzed. It was found that the optimum combined rate-power adaptation scheme achieves great performance (Jafar & Goldsmith, 2003), with rate adaptation being the main contribution factor. In this context, Yang and Hanzo (2004) showed that with adaptive SF, a 40% enhancement of total throughput is achieved (single-cell evaluation) without extra power consumption, quality degradation, and interference increase.

Uplink enhancement is already under consideration in 3GPP (TR 25.896 V.6.0.0), which introduces an enhanced uplink-dedicated channel (E-DCH) that is code multiplexed (multicode operation) with a different SF than normal dedicated channels, shorter Page 322  frame size (2 ms), uplink H-ARQ (Hybrid ARQ), and so forth. Results show a 50 to 70% cell-throughput enhancement compared to an R99 uplink.

Beam-Forming Techniques

Beam-forming techniques for both the downlink and uplink are based on the use of antenna arrays that focus the antenna beam to a specific region or user in order to improve the SIR. Several techniques have been studied (Li & Liu, 2003) and are under consideration for future use. The capacity enhancement achieved has the drawback of further installation and optimization costs for already operating networks. Thus such techniques will be rather engaged by new operators. It should be mentioned though that cost savings from the usage of common 2G to 3G antennas are lost.


The present status of commercial WCDMA and its future trends have been addressed. Entering the multimedia era forces operators to follow WCDMA enhancements as soon as possible in order to keep and expand their subscribers’ base. While first-launch implementation costs may not have been yet amortized, a future glance of 4G (4th Generation) and WLAN competition obligates operators to implement new WCDMA techniques and offer new services as soon as customers are ready to follow. Under these circumstances, the most cost-effective solutions will be selected. Among the different technologies described, transmit diversity schemes, HSDPA for downlink and link-adaptation methods and advanced receivers for uplink seem to be the next commercial step since the cost encumbers the operator. These techniques will strengthen the potential of multimedia provision and will provide adequate capacity and quality that will place an advantage of UMTS over 4G alternatives.

Fear and the Media - Fright Reactions to Individual Programs and Movies, Developmental Differences in the Effectiveness of Coping Strategies [next] [back] Fayruz (1933–) - PERSONAL HISTORY, BIOGRAPHICAL HIGHLIGHTS, PERSONAL CHRONOLOGY:, LEGACY OF A STAR, INFLUENCES AND CONTRIBUTIONS, THE WORLD’S PERSPECTIVE

User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or