IEEE Transactions on Quantum Engineering
A System-Level Design Framework for Rydberg-Atom Quantum Receivers: End-to-End Input-Output Modeling, Noise-Aware Sensitivity, and Communication Performance
Weichao Gao, Jinyun Wu, Peiyun Cheng, Bo Wu, Hao Zhang, Fuquan Li, Yinfa Zhang, Jian Yang
Abstract
Rydberg-atom electrometry promises a radiofrequency (RF) receiver front end with intrinsic self-calibration, all-optical readout, and freedom from conventional antenna constraints. However, existing studies largely address isolated physical effects—local susceptibility, spectral fitting, or a single observable–so a unified engineering description that links device physics to communication performance is still lacking. This work develops a system-level design framework that treats the Rydberg vapor cell as an RF receiver and connects the incident field to the electrical output along a single analytic chain. A closed-form static input–output map is obtained by combining a four-level continued-fraction susceptibility, Beer–Lambert propagation, and direct photodetection. Small-signal linearization yields a thirdorder complex-valued state-space model whose physical transfer function, because the photocurrent is a real projection of a complex coherence, must be assembled from both positive- and negative-frequency resolvent responses—a construction that has no counterpart in real-valued linear systems and is material for bandwidth prediction, which we show can otherwise be misestimated by up to a factor of two to three. A multisource noise model then defines a frequency-dependent minimum detectable field, exposing how static gain, spectral roll-off, and technical noise jointly bound sensitivity. The framework yields explicit reception limits for amplitude, frequency, and on–offkeyed modulation, unifies biased direct detection with superheterodyne operation, provides quantitative engineering design maps for bandwidth, sensitivity, and dynamic range, and is benchmarked against published87Rb and133Cs experiments, reproducing the measured calibration slope to within 2% and the reported bandwidth scale to order of magnitude.