Millimeter-wave Networking Fundamental Limits, Scalable Algorithms, and Design Insights

Abstract: The current demands for extremely high data rate wireless services and the spectrum scarcity at the sub-6 GHz bands are forcefully motivating the use of the millimeter-wave (mmWave) frequencies. The main characteristics of mmWave communications are severe attenuation, sparse-scattering environment, huge bandwidth, vulnerability to obstacles and antenna misalignment, massive beamforming, and possible noise-limited operation. These characteristics constitute a major difference with respect to legacy communication technologies, primarily designed for the sub-6 GHz bands, and are posing major theoretical design challenges that have not been sufficiently addressed so far. Motivated by these challenges, this doctoral thesis considers mmWave communications and investigates medium access control (MAC) layer design principles and performance analysis. Specifically, we focus on fundamental performance metrics, including coverage, fairness, robustness, throughput, and delay, which we address by three main research threads of increasing complexity.The first thread of the thesis analyzes the interference behavior in mmWave networks.We first propose a new index for assessing the accuracy of any interference model under any network scenario, which helps us develop a simple interference model of adequate accuracy. We then derive closed-form expressions for the throughput of mmWave ad hoc networks. The new analysis reveals that mmWave networks may exhibit a non-negligible transitional behavior from a noise-limited to an interference-limited behavior, depending on the system parameters such as density of transmitters, transmission power, and operating beamwidth. The second thread of this thesis builds on the previous one and addresses resource allocation in mmWave networks. For the short-term resource allocation, we establish a mathematical framework to investigate the impact of beam training (alignment) overhead on the network throughput. For the long-term resource allocation, we formulate a series of optimization problems that address relaying capability, frequent handovers, small multiuser interference, and load balancing. The third thread of this thesis extends the second one toward spectrum sharing in mmWave networks and characterizes the gains of beamforming and coordination in spectrum sharing via several optimization problems. We analyze these problems in the asymptotic regimes when the number of antennas becomes large and conclude that spectrum sharing with light on-demand coordination is feasible, especially at higher mmWave frequencies (for example, 73 GHz).The original analysis proposed in this thesis gives novel insights into many MAC layer issues such as resource allocation, interference management, random access, mobility management, and synchronization in future mmWave networks. The thesis also highlights that the design of mmWave networks poses open problems at the intersection of optimization and learning theories. Given the recent interest in the standardization of mmWave cellular networks and the highly sub-optimal nature of the existing standards for mmWave short-range networks, the results of this thesis may have the potential to substantially steer future standardizations.