Transmit Array Processing for CDMA Downlinks

It is expected that in the near future the demand for high data rates will dramatically increase especially in the downlink of mobile radio systems. This expectation is currently being reflected by the lively high speed downlink packet access (HSDPA) standardization efforts within 3GPP. CDMA mobile radio standards, e.g., the UTRA-FDD mode, are prepared for such requirements by means of orthogonal variable spreading factor (OVSF) codes allowing the coexistence of high and low data rate users. To achieve high data rates, low spreading factors have to be utilized, which goes along with a massive impact of interference and noise degrading the system performance. On the part of the transmitting base stations, high downlink data
rates may be enabled by utilizing adaptive multi-element transmit antennas. For the efficient operation of such antennas information about the downlink radio channels is required at the base stations. The 3G partial standard WCDMA designed for the Frequency Division Duplexing (FDD) radio frequency bands suffers from the inherent disadvantage that, due to the frequency gap between uplink and downlink, the results of the uplink channel estimation cannot be directly used as the channel information required for adjusting the multi-element transmit antennas. In the FDD case, two general basic approaches to obtain knowledge about the spatial properties of the downlink channels exist, namely either exploiting the spatial properties of the uplink channels, or feedback downlink channel information from the mobile stations to the supplying base stations. The first approach is based on the assumption that the directional properties and the attenuations of the respective radio propagation paths of the uplink channels and the corresponding downlink channels are equal. If this assumption holds, methods for an at least sub-optimum adjustment of the downlink transmit antenna weights based on uplink channel estimates exist, which do not rely on complex DOA estimation techniques. The comparison of the performances achievable by uplink channel estimation based techniques with those achievable if actual downlink channel information is available shows the superiority of techniques which directly exploit information about the spatial downlink channels. The only way to obtain information about the spatial downlink channels at the base station is via feedback from the mobile stations which leads to the development of efficient feedback schemes.

If among the mobile stations served by the same base station high and low data rate links have to coexist, mobile station specific minimum signal to noise and interference (SINR) constraints have to be fulfilled. Thus, transmit array processing and power control techniques should explicitly operate to support these demands. It is not possible to optimize transmit array processing and the adjustment of transmit powers in seperate steps. Iterative methods may be exploited to fulfill the mobile station specific SINR constraints and at the same time to keep the total transmit power as low as possible in order to reduce intercell interference.