Abstract
This thesis studies two distinct frontiers of quantum information processing: the fundamental physical limits of dynamical evolution and the practical realization of secure quantum communication networks.
In the area of applied cryptography, I address the critical challenge of implementation security in quantum key distribution (QKD). While, in principle, QKD offers unconditional security, real-world devices often contain imperfections that open “side-channels” for eavesdroppers. To close these loopholes, I propose and analyze fully passive sources that eliminate the need for active optical modulators, thereby removing a major class of source-side vulnerabilities. First, I develop a fully passive measurement-device-independent (MDI) QKD protocol. I construct a complete security framework involving a novel decoy-state analysis, and I show that secure communication is achievable over about 140km of standard optical fiber. Second, I extend this passive paradigm to the multi-user setting by introducing a fully passive conference key agreement (CKA) protocol based on twin-field interference. Using high-dimensional numerical integration and linear-programming techniques, I show that secure conference keys can be established among four users over channel losses up to roughly 28dB, corresponding to more than 130km of standard optical fiber, demonstrating that the protocol performs well even in such high-loss regimes.
In the area of fundamental dynamics, I study the quantum speed limit (QSL), which sets the ultimate bound on how fast a quantum system can evolve. I identify limitations in existing bounds regarding their tightness and universality. To resolve this, I derive a new family of QSLs based on representation-dependent weighted lpw-seminorms. This framework is universal, applicable to both closed and open system dynamics, as well as both time-dependent and time-independent evolutions, and extends beyond density matrices to general linear operators. I demonstrate that these new bounds are computationally efficient to optimize and provide good performance in various scenarios, including spontaneous emission and high-fidelity quantum gates.
Anyone interested is welcome to attend.