TY - GEN
T1 - Mutual information statistics of optimized LoS MIMO systems
AU - Matthaiou, Michail
AU - Pitarokoilis, Antonios
AU - Nossek, Josef A.
PY - 2010
Y1 - 2010
N2 - The presence of line-of-sight (LoS) components is typically considered as a hindrance for multiple-input multiple-output (MIMO) communications due to the limited amount of multipath scattering which, in turn, results in low spatial multiplexing gains. However, some recent investigations have questioned this common belief and demonstrated that by employing specifically designed antenna arrays at both the transmitter (Tx) and receiver (Rx), the mean channel matrix can become full-rank and, consequently, we can obtain high channel capacities even at high Ricean K-factors. In this paper, using the joint ordered eigenvalue probability density function (PDF) as a starting point, we derive analytical exact and asymptotic expressions for the mutual information (MI) statistics of these optimized LoS MIMO configurations. The proposed analytical formulae are given in a tractable determinant form and thus can be easily evaluated and efficiently programmed. The implications of the model parameters on MI statistics are also assessed with the match between the analytical curves and Monte-Carlo simulations being excellent.
AB - The presence of line-of-sight (LoS) components is typically considered as a hindrance for multiple-input multiple-output (MIMO) communications due to the limited amount of multipath scattering which, in turn, results in low spatial multiplexing gains. However, some recent investigations have questioned this common belief and demonstrated that by employing specifically designed antenna arrays at both the transmitter (Tx) and receiver (Rx), the mean channel matrix can become full-rank and, consequently, we can obtain high channel capacities even at high Ricean K-factors. In this paper, using the joint ordered eigenvalue probability density function (PDF) as a starting point, we derive analytical exact and asymptotic expressions for the mutual information (MI) statistics of these optimized LoS MIMO configurations. The proposed analytical formulae are given in a tractable determinant form and thus can be easily evaluated and efficiently programmed. The implications of the model parameters on MI statistics are also assessed with the match between the analytical curves and Monte-Carlo simulations being excellent.
UR - http://www.scopus.com/inward/record.url?scp=77955366812&partnerID=8YFLogxK
U2 - 10.1109/ICC.2010.5502245
DO - 10.1109/ICC.2010.5502245
M3 - Conference contribution
AN - SCOPUS:77955366812
SN - 9781424464043
T3 - IEEE International Conference on Communications
BT - 2010 IEEE International Conference on Communications, ICC 2010
T2 - 2010 IEEE International Conference on Communications, ICC 2010
Y2 - 23 May 2010 through 27 May 2010
ER -