Bidirectional adaptive optics architectures for optical communication through atmospheric turbulence

Abstract

Free-space optical (FSO) classical communication links can provide high data rates vital for successfully serving the world's exponentially growing demand for bandwidth, while FSO quantum key distribution (QKD) links allow information-theoretic rather than computational secure communication between two parties. Unlike fiber-optic classical communication and QKD links, FSO links can do so with minimal up-front investments in infrastructure. Setting aside absorption and scattering losses along the propagation path, optical links with a terrestrial terminal will still experience the deleterious effects of clear-weather turbulence, namely beam spread, beam wander, angle-of-arrival spread, and scintillation, which leads to low end-to-end power transfer from the transmitter to receiver. Decreases in the power transfer result in lower communication rates and may result in no secure-key rate for the loss-sensitive QKD communication protocols. Adaptive optics holds the best promise for mitigating, if not completely compensating for, these turbulence-induced degradations. Nevertheless, despite adaptive optics being a richly developed field, theoretical studies of adaptive optics systems have not fully exploited the reciprocal nature of propagation through atmospheric turbulence. It is known that applying ideal, full-wave adaptive optics at both the transmitter and receiver of a free-space optical link can guarantee scintillation-free power transmission when operation is deep in the near-field power transfer regime. Buoyed by the advent of enabling technologies like scalable Mach-Zehnder interferometer arrays and coherent receiver arrays, this thesis: (1) introduces both a full-wave and phase-only bidirectional adaptive optics (BDAO) protocol; (2) assesses the ergodic performances of FSO classical and QKD communication links utilizing these BDAO systems using both theoretical performance bounds as well as turbulence simulation results; and (3) provides an initial study of the noise and time-dynamics the BDAO protocol can tolerate while still achieving near-optimal classical communication or QKD rates.