Optimization Algorithm Design and Performance Analysis for Next-Generation Wireless Networks

Abstract

The proliferation of connected devices has led to very strict requirements for next-generation wireless networks, taking into consideration environmental as well as economic concerns. In particular, one of the primary goals in the design of fifth-generation (5G) wireless networks is to satisfy the extremely high data rate (traffic demand) of users with the minimum energy consumption. For this purpose, a new performance indicator, namely, energy efficiency (EE), has been proposed in the literature which is measured in bits/Joule and expresses the amount of information that can be reliably transmitted per unit of consumed energy. This Dissertation deals with the design of efficient optimization algorithms for next-generation wireless networks, including terrestrial as well as satellite communication systems. More specifically, the theory of sequential convex optimization (SCO) is applied to solve challenging optimization problems, such as the maximization of several EE-metrics, so as to develop energy-efficient power allocation strategies. SCO is a powerful mathematical tool that can be used to solve nonconvex optimization problems by solving a sequence of convex optimization problems. This method is theoretically guaranteed to converge for any initial feasible point and, under suitable constraint qualifications, achieves a stationary point (i.e., a point that satisfies the Karush-Kuhn-Tucker (KKT) conditions) of the original problem. Furthermore, we study some combinatorial optimization problems in satellite networks (SatNets), which are proven to be NP-hard. In particular, we focus on the optimum selection of ground stations (GSs) in SatNets with site diversity (SD), satisfying given availability requirements. SD technique is used to improve the availability of satellite systems by mitigating the atmospheric impairments, such as rain (for radio frequencies) and cloud coverage (for optical frequencies). Moreover, we present global optimization algorithms, based on the branch-and-bound (B&B) method and dynamic programming (DP), as well as a polynomial-time approximation algorithm with provable performance guarantee. Finally, we examine a load-sharing smart gateway diversity (LS-SGD) architecture in SatNets, which has been recently proposed in the literature. For this diversity scheme, we define the system outage probability (SOP) based on the Poisson binomial distribution (PBD) and taking into account the traffic demand as well as the gateway (GW) capacity. In addition, we present several methods for the exact and approximate calculation of SOP.