2026-05-23T22:29:31Zhttps://keep.lib.asu.edu/oai/request

oai:keep.lib.asu.edu:node-2026182025-10-29T23:09:56Zoai_pmh:alloai_pmh:repo_items

202618 https://hdl.handle.net/2286/R.2.N.202618 http://rightsstatements.org/vocab/InC/1.0/ All Rights Reserved 2025 178 pages Doctoral Dissertation Academic theses en Morais, Joao Alberto Janeiro Horta de Alkhateeb, Ahmed Chakrabarti, Chaitali Trichopoulos, Georgios C. Dasarathy, Gautam Arizona State University Partial requirement for: Ph.D., Arizona State University, 2025 Field of study: Electrical Engineering Emerging applications such as autonomous driving and augmented reality are pushing wireless networks to new limits. To meet increasing demands, networks adopt denser deployments, larger antenna arrays, and broader bandwidths at higher frequencies. However, these traditional scaling methods come at a cost—specifically, increased control signaling that reduces overall efficiency. A promising avenue to address this bottleneck lies in leveraging the strong correlation between wireless channel characteristics and device locations. This thesis explores that synergy: how positions can aid communication, how communications can enable localization, and how to bridge both through scalable, data-driven methods grounded in digital twins. It culminates in a study of dataset similarity, a key step to quantifying the realism of simulations and ensuring their usefulness in practice. This dissertation focuses on the synergy between locations and communications and on leveraging that synergy in real systems. Literature covers many use cases how location awareness can aid communication across all layers of the protocol stack. In that direction, this work contains the first real-world experimental study of using locations at the lower layers of the stack, namely for beam alignment, a key communication task in multi-antenna and mmWave systems. The inverse mapping, from communications to user locations, is explored less often due to the significant human effort associated with collecting communication data with location labels. This dissertation proposes a localization method that relies on scalable fingerprinting instead of human effort. This is achieved by leveraging digital twins to emulate real-world systems. In real systems, however, the complexity and unpredictability are currently incomparable to simulations, resulting in vastly different outcomes from theoretical and experimental studies. To study this discrepancy, this report also presents a real-world dataset and takes initial steps in comparing real and synthetic datasets. Electrical Engineering Artificial Intelligence Computer Science Artificial Intelligence Dataset similarity Digital Twins Localization Position-aware scheduling Wireless Communciations Toward AI-native Wireless Communication and Localization via Digital Twins