Applied Optimization Methods for Wireless Networks

Abstract

Written in a unique style, this book is a valuable resource for faculty, graduate students, and researchers in the communications and networking area whose work interfaces with optimization. It teaches you how various optimization methods can be applied to solve complex problems in wireless networks. Each chapter reviews a specific optimization method and then demonstrates how to apply the theory in practice through a detailed case study taken from state-of-the-art research. You will learn various tips and step-by-step instructions for developing optimization models, reformulations, and transformations, particularly in the context of cross-layer optimization problems in wireless networks involving flow routing (network layer), scheduling (link layer), and power control (physical layer). Throughout, a combination of techniques from both operations research and computer science disciplines provides a holistic treatment of optimization methods and their applications. Each chapter includes homework exercises, with PowerPoint slides and a solutions manual for instructors available online.

On user association and multiple access optimisation in 5G massive MIMO empowered ultra dense networks,

“…In particular, for a hotspot comprising M ANs serving K UEs, M K searches will be required, rendering the procedure impractical for network dimensions encountered in typical 3GPP-based small-cell cluster setups, involving tens of nodes. In the following two sections, we address the FDMA and SDMA user association problems by proposing two plausible exact reformulations, motivated by reformulation-linearisation concepts answered in applied optimisation [23,24].…”

A Survey of Network Lifetime Maximization Techniques in Wireless Sensor Networks

“…Therefore, a sensor barrier is usually formed by several connected sensors across the entire target region, which may feature a trip-wire-like structure to detect any potential crossings by intruders. An approximation algorithm that intelligently divides the search space into subareas, which are represented by "fictitious cost point" [154]. A near-optimal solution is obtained, where NL is guaranteed to be at least (1−ε) of the optimal NL provided that ε>0 and can be arbitrarily small.…”

A Scheduling Algorithm for MIMO DoF Allocation in Multi-Hop Networks

“…Although there already exist algorithms (e.g., Branch-and-Bound and sequential fixing [9]) and commercial optimization solvers (e.g., IBM CPLEX [10]) that can be used to solve the problem in Fig. 3, these algorithms/solvers are limited to the centralized environment.…”