Ultra wideband (UWB) systems are communication systems with a huge bandwidth (at least 500 MHz), operating in frequency bands that are already assigned to other services. To avoid interference with these legacy services, the transmit power per hertz of the UWB system is very low and must satisfy a mandatory predefined spectral mask. When UWB is combined with the multiple input multiple output (MIMO) principle, that uses multiple antennas at the transmitter and receiver, the UWB system becomes suitable for extremely high data rate short range communication, as e.g. wireless printing of digital pictures without the need of a PC, wireless televisions and monitors, etc. Reliable data detection in MIMO-UWB systems depends on the possibility to accurately estimate transmission parameters related to the channel and synchronization. Further, interference from legacy services operating in the same frequency band as the MIMO-UWB system must be suppressed. However, because of the huge bandwidth and low transmit power of the UWB signal, this parameter estimation and interference suppression is not trivial, and requires the development of advanced signal processing algorithms. An issue in the receiver design of a MIMO-UWB system is the absence of realistic channel models. Therefore, in this project proposal, we want to develop new realistic channel models for MIMO-UWB, based on measurements, which we can use in the derivation of new advanced signal processing algorithms.