On the connection between noncircularly-symmetric and noncentral fading modelsunivariate and multivariate analysis

  1. Moreno Pozas, Laureano
Dirigida por:
  1. Eduardo Martos Naya Director/a
  2. Francisco Javier López Martínez Director

Universidad de defensa: Universidad de Málaga

Fecha de defensa: 30 de junio de 2017

Tribunal:
  1. Marco Chiani Presidente/a
  2. Juan Manuel Romero Jerez Secretario/a
  3. Mohamed-Slim Alouini Vocal

Tipo: Tesis

Teseo: 479338 DIALNET lock_openRIUMA editor

Resumen

This thesis provides new statistical connections between noncircularly-symmetric central and circularly-symmetric noncentral underlying complex Gaussian models. This is particularly interesting since it facilitates the analysis of noncircularly-symmetric models, which are often underused despite their practical interest, since their analysis is more challenging. Although these statistical connections have a wide range of applications in different areas of univariate and multivariate analysis, this thesis is framed in the context of wireless communications, to jointly analyze noncentral and noncircularly-symmetric fading models. We provide an unified framework for the five classical univariate fading models, i.e. the one-sided Gaussian, Rayleigh, Nakagami-m, Nakagami-q and Rician, and their most popular generalizations, i.e the Rician shadowed, η-µ, κ-µ and κ-µ shadowed. Moreover, we present new simple results regarding the ergodic capacity of single-input single-output systems subject to κ-µ shadowed, κ-µ and η-µ fadings. With applications to multiple-input multiple-output communications, we are interested in matrices of the form W=XX^H (or W=X^H X), where X is a complex Gaussian matrix with unequal variance in the real and imaginary parts of its entries, i.e., X belongs to the noncircularly-symmetric Gaussian subclass. By establishing a novel connection with the well-known complex Wishart ensemble, we facilitate the statistical analysis of W and give new insights on the effects of such asymmetric variance profile.