DocumentCode :
2439719
Title :
Information-theoretic performance analysis of LMS MIMO communications
Author :
Alfano, Giuseppa ; De Maio, Antonio ; Tulino, Antonia M.
Author_Institution :
Corso Duca degli Abruzzi, Politec. di Torino, Torino, Italy
fYear :
2010
fDate :
Aug. 30 2010-Sept. 3 2010
Firstpage :
1
Lastpage :
5
Abstract :
Information-theoretic performance analysis of a MIMO communication over Land Mobile Satellite (LMS) channels, under ergodic and non-ergodic regimes, is performed. The capacity-achieving input covariance matrix, and the corresponding ergodic capacity, assuming perfect receive-side information but making different assumptions on the amount of channel knowledge at the transmitter, are derived. We obtain exact results, but for the case when perfect channel knowledge is assumed at both ends of the link, for which we provide an upper bound to the ergodic capacity. In the non-ergodic scenario, we compute the outage capacity in absence of power-control, and discuss the asymptotic Gaussianity of the mutual information, which strongly depends on the overall number of degrees of freedom available on the channel. Design guidelines for multiantenna LMS channels are gained studying the low Signal-to-Noise Ratio (SNR) behavior of the capacity, still under the assumption of absence of knowledge of the channel matrix (or its statistics) at the transmitter. The results are illustrated through several examples, aimed at assessing the impact on the performance of the diversity order and/or the Line-of-Sight (LOS) fluctuations.
Keywords :
MIMO communication; covariance matrices; information theory; MIMO communication; channel knowledge; channel matrix; covariance matrix; information theory performance analysis; land mobile satellite channel; line of sight fluctuation; multiantenna LMS channel; power control; signal to noise ratio; Covariance matrix; Least squares approximation; MIMO; Mutual information; Receiving antennas; Signal to noise ratio; Transmitters;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Theory Workshop (ITW), 2010 IEEE
Conference_Location :
Dublin
Print_ISBN :
978-1-4244-8262-7
Electronic_ISBN :
978-1-4244-8263-4
Type :
conf
DOI :
10.1109/CIG.2010.5592839
Filename :
5592839
Link To Document :
بازگشت