Analysis of Millimeter Wave Wireless Relay Networks

Abstract

Fifth generation (5G) wireless networks must handle massive data rates given the recent and future demand growth. A solution is the use of millimeter wave (mmW) bands (20-100 GHz) which can leverage massive bandwidths (up to 5 GHz). However, mmW signals experience high path loss, directivity and blockages. Moreover, node locations are becoming increasingly random due to deployment constraints which limits the performance of these networks. To overcome these challenges, this thesis makes several contributions on the deployment of randomly located mmW relays to enhance the coverage and rate. First, an analytical framework is developed for the coverage and rate of a decode-and-forward relay network with a best relay selection by incorporating mmW channel impairments and by using stochastic geometry technique to model the relay locations. Second, a multi-hop network is analyzed in noise-limited and interference-limited regimes to characterize the coverage, error rate, and ergodic capacity by developing the end-to-end signal-to-noise ratio distribution. Third, to improve the relaying spectral efficiency and coverage, two-way relays with two users exchanging their message is analyzed by selecting a best relay from a set of randomly located nodes. Finally, coverage is analyzed when non-orthogonal multiple access is utilized in mmW relaying for several relay selection strategies. This thesis demonstrates that these relaying techniques improve coverage and rate significantly when appropriate relay selection is performed.