Abstract: In this paper we show that geometric optimization is the most efficient way to reach maximum channel capacity when the channel is characterized by a powerful Line-Of-Sight signal component but it will result in larger antenna spacings than they are found customary. To have a quantitative criterion for comparison we firstly derive boundaries for the channel capacity when considering a pure Line-Of-Sight signal component. For finding the upper and lower bound we take into account signal pre-processing strategies in conjunction with systematic power allocation schemes. Having detected these bounds we calculate the capacity in the case of a linear antenna array spaced half-wavelength at both the transmitter as well as the receiver as it is commonly chosen in mobile communications. The capacity turns out adopting values close to the lower bound even if it is optimized best by means of signal processing and power allocation. On the contrary we present an examplary geometrically optimized set of antenna elements, which is dedicated to reach the upper bound of the channel capacity.
Keywords: Multiple Input - Multiple Output, MIMO, wireless communications, capacity, Line-Of-Sight, correlated channels, keyhole channels, parallel channels, non-frequency selective channels.