Performance of Downlink Spatial-Multiplexing in Realistic Urban Multi-User MIMO Channels

In this paper the performance of user specific beamforming, waterfilling, block diagonalization, and successive optimization is compared to the performance of a conventional SISO reference system in a single cell with a sectorized BS and several MTs. Whereas in the SISO reference system as well as in the beamforming and waterfilling algorithm the MTs are multiplexed by a conventional multiplexing method (TDMA, FDMA, CDMA) the two multi-user MIMO algorithms allow users to share the same spectrum by multiplexing them in space. Several multi-user scenarios are generated by means of a realistic urban channel model. The channel model incorporates a model of the city of Karlsruhe, a mobility model as well as a ray-tracing based wave propagation model. Simulation results show clearly the potential of downlink spatial-multiplexing to improve the system capacity, to reduce the exposure (power density) and to reduce the transmit power and therefore the operating costs in future cellular mobile radio systems.