Ivrla?, Michel T.; Nossek, Josef A. (Institute for Circuit Theory and Signal Processing, Technische Universität München, 80333 Munich, Germany)
Electromagnetics provides the ground for a physical theory of communications, while information theory and signal theory approach the problem from a purely mathematical point of view. Nevertheless, the latter theories frequently do refer to physical terms, such as: energy, power, noise, antennas, or waves. It is strange enough, that, at present time, there is no provision being made such that the usage of such terms in information theory is at least consistent with the governing physics. More often than not, this results in less than optimum signal processing solutions and does not contribute to a complete understanding of communication systems. Circuit theoretic channel models can help to bridge the gap between the physics of electromagnetics, and the mathematical world of information theory. The multiport concept makes sure that important physical concepts like energy, power, or noise are captured correctly, and terms such as antennas or waves are applied consistently with their physical meanings in information theory and signal processing. We suggest how to make circuit theoretic channel models and apply them to wireless communication systems which uses multiple antennas at both ends of the link. We thereby show that, in contrast to common belief, arrays of closely spaced antennas actually do support the bandwidth- and power-efficient multi-stream transmission.