In cellular communication systems, one of the prominent impairments due to the wireless channel is the multipath fading. In this thesis, we study the multipath effects, particularly in the so-called cascaded fading channels. These channel models are useful when characterizing the indoor signal propagation, time-reversal transmission, and the rank deficiency due to insufficient scattering in multiantenna communications. Thesis focuses on the statistical performance metrics of wireless systems. More precisely, in single-antenna transmission, we analyze the received signal distribution via moment-based approximation in presence of correlated cascaded channels. Using these results, the impact of channel correlation and number of scatterers are investigated for the cascaded Nakagami-m channels. Furthermore, the signal distribution of the cascaded Rician channel is used to construct a test statistics for a blind time-reversal detector. In the presence of a point target embedded in the multipath scattering, we show by Monte Carlo simulation that the proposed detector outperforms the existing detectors. In multi-antenna communications framework we consider information theoretic limits for the cascaded Rayleigh MIMO channels. In the ergodic channel, we obtain a lower bound for the ergodic mutual information and deduce a rate scaling law using a recent result from the random matrix theory. For the non-ergodic channel, we study the outage probability by establishing the central limit theorem for the linear spectral statistics of the channel matrices. This result also motivates us to construct a simple approximation for the fundamental tradeoff between diversity gain and multiplexing gain of the MIMO channel at realistic SNR levels.
|Translated title of the contribution||Statistical Analysis of Cascaded Multipath Fading Channels|
|Publication status||Published - 2015|
|MoE publication type||G5 Doctoral dissertation (article)|
- cascaded fading channel
- free probability theory
- random matrix theory
- wireless communications