Resumen de Closed form analysis of poisson cellular networks: a stochastic geometry approach
Ultra dense networks (UDNs) allow for efficient spatial reuse of the spectrum, giving rise to substantial capacity and power gains. In order to exploit those gains, tractable mathematical models need to be derived, allowing for the analysis and optimization of the network operation. In this course, stochastic geometry has emerged as a powerful tool for large-scale analysis and modeling of wireless cellular networks. In particular, the employment of stochastic geometry has been proven instrumental for the characterization of the network performance and for providing significant insights into network densification. Fundamental issues, however, remain open in order to use stochastic geometry tools for the optimization of wireless networks, with the biggest challenge being the lack of tractable closed form expressions for the derived figures of merit.
To this end, the present thesis revisits stochastic geometry and provides a novel stochastic geometry framework with a twofold contribution. The first part of the thesis focuses on the derivation of simple, albeit accurate closed form approximations for the ergodic rate of Poisson cellular networks under a noise limited, an interference limited and a general case scenario. The ergodic rate constitutes the most sensible figure of merit for characterizing the system performance, but due to the inherent intractability of the available stochastic geometry frameworks, had not been formulated in closed form hitherto. To demonstrate the potential of the aforementioned tractable expressions with respect to network optimization, the present thesis proposes a flexible connectivity paradigm and employs part of the developed expressions to optimize the network connectivity. The proposed flexible connectivity paradigm exploits the downlink uplink decoupling (DUDe) configuration, which is a promising framework providing substantial capacity and outage gains in UDNs and introduces the DUDe connectivity gains into the 5G era and beyond.
Subsequently, the last part of the thesis provides an analytical formulation of the probability density function (PDF) of the aggregate inter-cell interference in Poisson cellular networks. The introduced PDF is an accurate approximation of the exact PDF that could not be analytically formulated hitherto, even though it constituted a crucial tool for the analysis and optimization of cellular networks. The lack of an analytical expression for the PDF of the interference in Poisson cellular networks had imposed the use of intricate formulas, in order to derive sensible figures of merit by employing only the moment generating function (MGF). Hence, the present thesis introduces an innovative framework able to simplify the analysis of Poisson cellular networks to a great extent, while addressing fundamental issues related to network optimization and design.
