Sep 9 (Tues) @ 2:00PM: "Optimization of Emerging Relaying Paradigms in Next-Generation Communication Networks," Winston Hurst, ECE PhD Defense

Zoom Meeting: https://ucsb.zoom.us/j/88579819067?pwd=4ajB7I2IC8XscRDssQUU2YdkDSKHPn.1
Research Areas: Communications & Signal Processing, Control Systems
Research Keywords: Control, Wireless Communication, Optimization,  Game Theory

Abstract

Next-generation communication systems will leverage emerging paradigms to improve capacity, latency, reliability, and coverage, while enabling new applications such as autonomous vehicles and AR/VR. In this context, this talk considers two novel relay systems that enable dynamic and flexible topologies. First, we examine autonomous mobile relays, which can provide on-demand network capacity to replace or augment terrestrial infrastructure. We study the problem of minimizing the average waiting time of data at a source node before it is relayed to the corresponding destination. Modeling multiple source–destination pairs as a generalized polling system, we design optimized relay trajectories using both a mixed-integer second-order cone formulation and a novel reinforcement learning approach.

We then investigate the use of wireless power transfer (WPT) to enable non-cooperative relaying (NCR), thereby reducing the infrastructure otherwise required to overcome the challenges of mmWave propagation. We propose an auction-based protocol in which energy payments, delivered via WPT, compensate end users for providing relay service. We analyze both Vickrey and Myerson auctions, rigorously characterizing fading conditions under which the latter is regular—a property ensuring computational tractability. Analytical results, corroborated by extensive simulations, show that NCR performance approaches that of a cooperative baseline under both auction schemes. Building on this foundation, we extend NCR to the case of an autonomous vehicle (AV) traversing a known path while collecting data. With limited onboard buffer capacity, the AV must offload data efficiently to avoid overflow. We address this in two steps. First, we design a novel multi-attribute mechanism based on the second-preferred offer (SPO) auction, which reduces the auction outcome to a function of a controllable weighting factor and a random variable. Second, we develop control strategies for the weighting factor using barrier functions and model predictive control to prevent buffer overflow while maintaining energy efficiency. Overall, these results highlight how relay-aided communication, supported by mobile relays, wireless power transfer, and auction-based coordination, can provide dynamic and energy-efficient solutions for next-generation networks.

Bio

Winston Hurst received the B.S. degree in computer science from Brigham Young University, Provo, UT, USA, in 2016, and the M.S. degree in electrical engineering from the University of California, Santa Barbara, CA, USA, in 2021. He is currently pursuing the Ph.D. degree in electrical and computer engineering at the University of California, Santa Barbara. His research interests include game theory and incentive design in wireless networks, mobility-enabled communication systems, and control and optimization in stochastic environments.

Submitted By: Winston Hurst <winstonhurst@ucsb.edu>