Semi-Blind Signal Detection with Adaptive Antennas for CDMA

The steadily increasing penetration of mobile communication services as well as the upcoming introduction of multimedia services (like there will be in UMTS) raises the requirements which the mobile communication systems will have to fulfil. A solution to this problem are adaptive antennas. They consist of an antenna array, a transmit/receive structure and a signal processing unit. Because of the directional nature of the mobile radio channel adaptive antennas can separate different users transmitting over different channels. To achieve this goal, the signals of the antenna elements are combined in such a way, that the signal of the user of interest is enhanced, while the other user signals are suppressed. This can be done by so called semi-blind algorithms. These algorithms do not need information about the geometrical structure of the antenna array nor calibrated transmit/receive-structures, which simplifies the construction of the adaptive antenna. They use instead structural properties of the sent signal as well as training symbols which are sent by most of today`s digital mobile communication systems. The purpose of this research project is to investigate such semi-blind algorithms for Code Division Multiple Access (CDMA) systems. In the beginning we will focus on computationally efficient algorithms based on previous work that was done by Juha Laurila with GSM in mind (DILFAST). This algorithm needs surprisingly low computational power so that it can be used by real-time systems. We want to exploit this great potential for CDMA signals. Afterwards we will move on to incorporate the available information of the other active users of the cell into the algorithm and finally we will also have a look on subspace based algorithms. All investigated algorithms will be compared by means of their bit-error-rate over the signal to noise ratio as well as their computational efficiency for different numbers of interfering users. Additionally we want to investigate non-conventional antenna element distributions in order to improve on pure "beamforming" or "diversity" structures of conventional smart antennas.