Analysis and comparison of optimal predictive transmitter selection and combining diversity for DECT

DECT (Digital European Cordless Telecommunications) is the European standard for digital cordless telecommunications systems. It operates in a pico-cellular, unregulated environment. The transmit power of DECT has been confined to reduce inter-system interference. Thus range improving diversity measures have received widespread attention. For form-factor reasons they are confined to the base-station. This is possible, because DECT uses time division duplex. The best receive antenna is selected for transmission in the next TX time slot (transmitter selection diversity). This approach has two drawbacks: (i) the diversity gain is very sensitive to the mobile speed, (ii) for higher order diversity the gain is unsatisfactory. The use of a predictor has been proposed to increase the robustness regarding mobile speed. The scheme is very sensitive however to quantization noise in the A/D conversion. In this paper two TX antenna diversity schemes are investigated and compared which solve the drawbacks of the conventional approaches: (i) optimal predictive selection diversity and (ii) optimal predictive combining diversity. Analytical expressions are given for the average probability of error. Numerical results are presented, which highlight the 2 advantages of either approach in a DECT framework. Both schemes are robust to mobile speed and quantization noise. At 10 km/h they outperform known schemes by 5 and 8 dB respectively (3 antennas, target bit error rate 10^-3)