Asymptotic ergodic capacity of multidimensional vector-sensor array MIMO channels

Abstract

We analyze asymptotic ergodic capacity of multidimensional vector-sensor array MIMO (PMD-MIMO) channels established by the use of dual-polarized antennas in the form of 1D, 2D and/or 3D MIMO arrays. Based on the identification of the decomposition of PMD-MIMO channels into multiple independently-fading and scaled classical MIMO channels in parallel, we consequently derive corresponding asymptotic ergodic capacities analytically via tools out of free probability theory. The analysis of derived asymptotic ergodic capacity expressions in terms of antenna locus aspect ratio beta, average symbol SNR per antenna gamma3 and cross-polar discrimination XPD as well as comparison with asymptotic ergodic capacity of classical MIMO channels present important gains in using compact multidimensional vector-sensor array MIMO systems in asymptotic regimes.