Co-optimization of linear gain and dynamic range for atomic superheterodyne receivers based on homodyne readout

Abstract Rydberg-atom-based superheterodyne receivers integrate self-calibration, high sensitivity, a wide operational frequency range, and phase/frequency resolved detection capabilities, demonstrating broad application prospects as next-generation microwave receivers. Linear gain and linear dynamic range (LDR) are critical metrics for assessing receiver sensitivity and demodulation fidelity, respectively. We numerically solve the four-level master equation and then employ particle swarm optimization (PSO) algorithm to co-optimize linear gain and LDR in atomic superheterodyne receivers based on balanced homodyne detection. Further, we systematically account for dominant dephasing mechanisms in the simulation, encompassing spontaneous decay, transit dephasing, collision dephasing, laser linewidth dephasing, and Doppler averaging. Homodyne readout utilizes both the real and imaginary parts of polarizability for sensing. In the case of the photon shot noise limit, its signal-to-noise ratio (SNR) expression resembles that of direct optical-intensity readout. However, the inherent coherent subtraction operation in homodyne detection significantly suppresses common-mode noise, while appropriately increasing the reference beam power enhances the gain in practical experiments. Indeed, this co-optimization problem, characterized by a high-dimensional variable space, two objectives, and non-convexity, is well-suited for solution by PSO. In addition, probe and coupling detuning contribute equivalently to polarizability and compensate for each other owing to Doppler averaging, thereby reducing the optimization variable space by one. By adopting a product form of linear gain and LDR as the fitness function, the PSO achieves rapid convergence. Here, the effectiveness of the PSO results is verified via the total harmonic distortion (THD). The relative error-based LDR calculation method we proposed efficiently measures receiver response linearity with consuming fewer computational resources. This research is expected to offer valuable insights into enhancing the performance of Rydberg-atom-based superheterodyne receivers.