Cost efficient representation and signaling of channel information in MIMO systems

Radio communications systems, which utilize antenna arrangements consisting of several transmit and receive antennas, have inspired theorists and are now on the verge of practical application, for instance in future mobile radio systems. In order to fully benefit from the potential offered by such MIMO (multiple input multiple output) systems, the knowledge of the radio channels should be available and considered in the signal processing algorithms both at the transmitter and the receiver. At the receiver this knowledge can be gained by state of the art training sequence based, semi-blind or blind channel estimation. However, at the transmitter the channel knowledge is not readily available and has to be imparted from the receiver by signaling the relevant information back to the transmitter. Unfortunately, this back signaling ties up transmission resources. Therefore, a cost efficient signaling concept for this purpose is proposed, which relies on subspace methods. Utilizing the proposed signaling concept allows an efficient exploitation of the feedback resources. A quantitative example is discussed to show how the concept can be applied and how it performs in different scenarios.