As wireless communications continue to permeate many aspects of human life and technology, future generations of communication networks are expected to become increasingly heterogeneous due to an explosion of the number of different types of user devices, a diverse set of available air interfaces, and a large variety of choices for the architecture of the network core. This heterogeneity, coupled with increasingly strict demands on the communication rate, latency, and fidelity demanded by a growing list of services delivered using wireless technologies, requires optimizations across the entire networking stack. Our contribution to this effort considers three key aspects of modern communication systems: First, we present a set of new techniques for multiple-input, multi-output beam alignment specifically suited for unfavorable signal-to-noise ratio regimes like the ones encountered in beamformed millimeter-wave wireless communication links. Second, we present a computationally efficient estimation algorithm for a specific class of aeronautical channels, which applies to systems designed to extend wireless coverage and communication capacity using unmanned aerial vehicles. Third, we present a new class of multi-hop relaying schemes designed to minimize communication latency with applications in the emerging domain of ultra-reliable and low-latency communications. Each of the three problem areas covered in this work is motivated by the demands of a future generation of wireless communication networks and we develop theoretical and/or numerical results outperforming the state of the art.