Search

The WCDMA specification is constantly evolving. In this paper we present results from measurements on a test bed, based on the WCDMA enhanced uplink (UL) concept, which is being introduced in the 6th release of the WCDMA specification. The enhanced UL test bed has been implemented on top of a commercial WCDMA release 99 cellular system and integrated with an existing high speed downlink packet access (HSDPA) test bed. Measurements have been performed over a coaxial network that includes a fading channel emulator. Three transport formats have been measured: 336, 672 and 1344 kbit/s. The measured performance is very close to simulated performance at moderate block error rates (BLER). For high signal-to-noise ratios an error floor causes some degradation.

A simple model for rate and modulation "interpolation" of 3GPP turbo code performance of the first transmission is studied. To ease the application of the information based quality model, the paper proposes an approximation for the information expression, and proposes a polynomial to calculate the needed information for a given coding rate. The proposed approximation and fitting polynomial are applied to the link adaptation in HSDPA to select the best transport format and resource combination (TFRC). With this proposal, simulations for only a few TFRC are needed to construct the model, and the performance of other TFRC are predicted by this model. The model is applicable for different modulation schemes, coding rates and mobile speeds.

The paper studies closed loop (CL) transmit diversity in combination with high speed downlink packet access (HSDPA). A potential problem with this combination is the reduced interference predictability which affects scheduling and link adaptation. To study this, the orthogonality factor (OF), which is used to model the intra-cell interference for single antenna transmission and reception, is augmented by an orthogonality matrix (OM), which is used for dynamic system level evaluations. The results indicate that the reduced predictability may decrease system performance in terms of cell throughput by approximately one third for a system employing multiuser diversity in a dispersive environment. However, by scheduling multiple users in a certain proposed way, interference fluctuations may be reduced, and results show that the performance may be improved. A simple spatial channel model is used so that the impact of fading correlation is considered as well.

In WCDMA systems, in particular with the introduction of high-speed multimedia services, the demand for good indoor coverage will become important. Overlaying macro cells can provide a sufficient level of general indoor coverage for most of the cases. However, in order to be able to provide WCDMA indoor coverage and capacity within traffic hot spots, such as airports, shopping malls and large office buildings, the deployment of dedicated in-building systems may be required. In this paper, the gain of deploying dedicated in-building systems in terms of HSDPA indoor coverage and capacity is studied; first, with a simple analytical reasoning and after that, with advanced system simulations. The results show that the HSDPA performance can be improved considerably within the indoor hotspot area by deploying a dedicated in-building system. The results also demonstrate how a dedicated in-building system can off-load the overlaying macro cell, and as a result of that, improve the grade of service considerably for the co-existing non-HSDPA users.

The high-speed data services, up to about 14.4 Mbps, for the 3G networks is supported by the high-speed downlink packet access (HSDPA) concept. Adaptive modulation and coding (AMC), multicode operation and hybrid automatic repeat request (HARQ) are the enabling technologies used for HSDPA. The next generation mobile communications systems is anticipated to support even higher data rates up to and exceeding 100 Mbps. MC-CDMA and DS-CDMA, both using frequency-domain equalization, are considered the most promising wireless access schemes. In this paper, we compare the throughput performance of MCand DS-CDMA when packet data are transmitted as in HSDPA with AMC and HARQ using multiple codes.