TABLE OF CONTENTS Appropriate modeling of multiple-input/multiple-output (MIMO) channels is naturally the first step in the study of multiantenna systems. Before starting the analysis, it is important to clarify that, even though MIMO channel systems are usually intended to be multiantenna systems, this is not always the case. Multiple antennas are necessary only to create multiple spatial channels, but other dimensions can be exploited as well, such as time, frequency, and polarization. For example, multiple temporal channels can be implemented sending the information bits through different time slots, as in time division multiple access (TDMA) systems; frequency channels can be induced by sending bits through different (sub-) carriers, as in frequency division multiple access (FDMA) systems. Moreover, since in wireless communications the transmitted signals are carried by electromagnetic fields, which are vectors characterized by a certain polarization, we can also associate different bits to different polarization states, so as to create multiple polarization channels. One further possibility arises when planar antennas are used. Planar antennas (like drums) are in fact characterized by multiple modes that could be excited or sounded independently of each other. Hence, space, time, frequency, polarization, and even vibration modes, are all examples of physical channels that can be used to transmit information. We can thus see a MIMO system as the instrument to convey information through multiple channels. The nature of the channels can be very diverse. What is important is the multiplicity of the channels. This is indeed a possibility not far...
