Calibration-Free Rydberg Atomic Receiver

• The paper introduces a calibration‐free receiver that uses atomic parameters like EIT, Autler–Townes splitting, and Stark shifts to achieve SI‐traceable RF field detection.
• The methodology relies on ladder-type Rydberg EIT and diverse atomic level schemes to perform demodulation, heterodyne reception, and multiband recovery in a single vapor-cell platform.
• Engineered architectures such as cavity coupling, self-heterodyne comb readout, and adaptive LO tracking enhance sensitivity and dynamic range for practical RF communication.

A calibration-free Rydberg atomic receiver is a radio-frequency or microwave receiver in which an incident electric field is converted into an optical observable by a vapor of Rydberg atoms, and the field amplitude is inferred from intrinsic atomic parameters rather than from an externally calibrated antenna factor. In this class of devices, electromagnetically induced transparency, Autler–Townes splitting, and Stark shifts provide an SI-traceable link between the received field and probe-laser transmission, enabling field sensing, heterodyne reception, demodulation, and, in some architectures, phase, polarization, or multi-band recovery within a single vapor-cell platform (Xie et al., 1 Jun 2026, Dixon et al., 2022, Chen et al., 15 Jan 2025).

1. Quantum-optical basis of calibration-free reception

The standard operating principle is a ladder-type Rydberg-EIT configuration. A weak probe laser drives a ground-to-intermediate transition, a strong coupling laser drives an intermediate-to-Rydberg transition, and an RF or microwave field couples adjacent Rydberg levels. In the resonant RF regime, the relevant relation is

$$\Omega_{\rm RF}=\frac{d_{\rm RF}E_{\rm RF}}{\hbar},\qquad \Delta f_{\rm AT}=\frac{\Omega_{\rm RF}}{2\pi},$$

so that

$$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$

Because the dipole matrix element is taken from atomic theory, the measured Autler–Townes splitting yields the local electric field without external RF-field calibration (Dixon et al., 2022).

The same principle appears in S-band cesium receivers, where the microwave field drives the $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$ Rydberg transition and the EIT peak splits in proportion to the field amplitude. In heterodyne or superhet operation, a strong LO and a weak signal detuned by $\delta$ generate a beat note in the transmitted probe power,

$$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$

with $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$, so amplitude and phase information are encoded directly in the optical readout (Xie et al., 1 Jun 2026).

At sub-MHz frequencies, the mechanism changes because direct Autler–Townes readout becomes ineffective. The field is then modeled as

$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$

which produces a quadratic Stark shift. The measurable splittings

$$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$

permit elimination of the unknown bias field and direct recovery of the AC amplitude,

$$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$

This removes dependence on electrode spacing or field-amplitude references and preserves calibration-free operation through intrinsic polarizability (Chen et al., 15 Jan 2025).

In analytical communication models, the received RF fields imprint amplitudes and phases onto the steady-state probe coherence $\rho_{2,1}$, and the optical output is written as

$$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$0

The receiver output is therefore an optical measurement of atomic susceptibility rather than an electrical voltage at a matched antenna terminal (Shyamal et al., 13 Apr 2026).

2. Atomic level schemes and receiver architectures

A widely used implementation is the four-level cesium ladder. In the S-band cavity-enhanced receiver, the levels are $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$1, $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$2, $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$3, and $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$4. An 852 nm probe and a 509 nm coupling beam counter-propagate through a room-temperature cylindrical Cs vapor cell of length 100 mm and diameter 30 mm. The cavity-based architecture places the cell at the electric-field antinode of a rectangular copper resonator operating in the fundamental $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$5 mode near $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$6 (Xie et al., 1 Jun 2026).

A second cesium implementation replaces laser scanning with a self-heterodyne optical frequency comb. There, a probe comb centered at 852 nm is generated with an electro-optic modulator and arbitrary waveform generator, producing approximately $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$7 evenly spaced teeth over approximately $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$8, while a single-frequency 509 nm coupling laser addresses the Rydberg launch state. The transmitted comb is then beat with a local oscillator derived from the same probe laser on a fast photodiode, so the transmission of each comb tooth is observed in parallel (Dixon et al., 2022).

Architectural generalization beyond the four-level ladder is explicit in the six-level hybrid Rydberg atomic quantum receiver. That system combines a probe transition $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$9, a coupling transition $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$0, and four RF-driven Rydberg-manifold transitions $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$1 through $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$2. Pathways $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$3–$\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$4 form a cascaded chain and $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$5 forms a parallel branch, yielding

$\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$6

simultaneous RF channels within one six-level system (Shyamal et al., 13 Apr 2026).

Other architectures are optimized for distinct functions. A warm $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$7 receiver near the $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$8 Wi-Fi band uses a three-photon ladder with 780 nm, 776 nm, and 1268 nm lasers and a $\lvert 3\rangle\!\leftrightarrow\!\lvert 4\rangle$9 microwave transition near $\delta$0 for heterodyne QAM reception (Nowosielski et al., 2024). A DC-Stark-enabled receiver uses a four-state loop in which a static bias mixes the upper Rydberg pair, activates an otherwise absent optical pathway, and enables phase-sensitive recovery of a single RF signal without injecting an RF LO into the atoms (Katkov et al., 31 Mar 2026). This suggests that calibration-free operation is not tied to one particular level diagram, but to the atomic transduction principle.

3. Readout modalities and communication functions

The earliest communication demonstrations emphasized direct demodulation inside the atomic medium. In an atom-based receiver for AM and FM microwave communication, the coupling laser is fixed on a steep flank of the RF-dressed EIT resonance so that instantaneous probe transmission responds to either the field amplitude or the field detuning. The reported 3 dB baseband bandwidth is approximately $\delta$1 for both AM and FM, the dynamic range is $\delta$2–$\delta$3 at $\delta$4, and the second harmonic in AM appears at least $\delta$5 below the fundamental. Carrier frequencies from $\delta$6 to Q-band were received by retuning the coupling laser to different Rydberg transitions (Anderson et al., 2018).

Heterodyne reception substantially extends the modulation repertoire. In the S-band superhet configuration, a strong LO and a weak signal are combined before entering the Rydberg medium, and the beat note is recovered from the photodiode output (Xie et al., 1 Jun 2026). In the warm-vapor QAM receiver, two dipole antennas illuminate the cell with an LO and a quadrature-modulated signal; the atoms mix the two fields, the avalanche photodiode detects the intermediate frequency, and a digital demodulator recovers $\delta$7 and $\delta$8. The heterodyne amplitude was measured up to approximately $\delta$9 before falling to the shot-noise floor, and for a single 4-QAM channel at $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$0 the coded capacity was reported as approximately $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$1 with $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$2 (Nowosielski et al., 2024).

Low-frequency communication around $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$3 has been demonstrated using BPSK, OOK, and 2FSK in a cesium cell with internal parallel electrodes. The reported error vector magnitude is $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$4 at $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$5, $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$6 at $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$7, and $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$8 at $$P_{\rm out}(t)=P_0+\kappa\,\Omega_{\rm SIG}\cos(2\pi\delta t+\phi_{\rm SIG}),$$9. A high-fidelity digital color image transmission produced a peak signal-to-noise ratio of $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$0 (Xie et al., 2024). In the sub-MHz Stark-modulation receiver, experiments at $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$1 established calibration-free field retrieval from cycle-averaged spectra rather than from conventional Autler–Townes splitting (Chen et al., 15 Jan 2025).

For high-dynamic scenarios, adaptive LO tracking has been proposed to keep the intermediate frequency inside the narrow atomic response window. The architecture uses a cross-product automatic frequency control loop with digital PI filtering and dual correction paths through an NCO and a VCO. In simulation, under a linear Doppler rate of $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$2, the adaptive architecture keeps the IF tightly centered at $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$3 with residual tracking error not exceeding approximately $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$4, whereas a fixed-LO architecture drifts beyond the atomic instantaneous bandwidth in less than $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$5 (Xiang et al., 9 Jul 2026).

4. Sensitivity engineering and reported performance

A central systems problem is inefficient coupling between free-space fields and the atoms. The S-band cavity receiver addresses this by combining a horn antenna with a resonant copper cavity of internal dimensions $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$6. The simulated unloaded quality factor is approximately $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$7, the measured unloaded quality factor is approximately $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$8, the measured $\Omega_{\rm SIG}=\mu_{34}E_{\rm SIG}/\hbar$9 is approximately $E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$0, and with antenna/cable coupling $E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$1 the loaded quality factor is approximately $E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$2. In this geometry the cavity amplifies the incident field at the atomic sample, and the horn collects the free-space signal over a $E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$3 path (Xie et al., 1 Jun 2026).

Configuration	Calibration coefficient	Reported sensitivity
Bare cell (Mode A)	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$4	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$5
Direct cavity injection (Mode B)	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$6	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$7
Antenna-coupled cavity (Mode C)	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$8	$E(t)=E_{\rm DC}+E_{\rm AC}\cos(2\pi f_{\rm AC}t),$9 with antenna gain; $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$0 without antenna gain

In that S-band study, the best performance is obtained in the antenna-coupled cavity mode, yielding an improvement of approximately $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$1 over the optimized bare vapor-cell configuration. The noise-equivalent field is defined as the field amplitude that gives $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$2 in a $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$3 bandwidth, and is computed from the measured SNR, signal power, resolution bandwidth, and calibration factor (Xie et al., 1 Jun 2026).

Other reported sensitivity figures depend strongly on regime and readout. The self-heterodyne comb electrometer resolved EIT linewidths below $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$4 and reported a minimum detectable RF field of approximately $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$5 with sensitivity approximately $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$6 (Dixon et al., 2022). The warm $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$7 heterodyne QAM receiver reported $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$8 in $$\Delta f_1=\frac{1}{h}\,\tfrac12\,\alpha_iE_{\rm DC}^2,\qquad \Delta f_2=\frac{1}{h}\,2\,\alpha_iE_{\rm DC}E_{\rm AC}$$9 and a noise-equivalent field $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$0 (Nowosielski et al., 2024). The sub-MHz receiver reported a minimum detectable field of $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$1 at $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$2 (Chen et al., 15 Jan 2025). More generally, the hybrid multi-band study states that sensitivity in Rydberg receivers can reach the $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$3 level experimentally, with theoretical limits in the $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$4 range (Shyamal et al., 13 Apr 2026).

5. Bandwidth limits, noise sources, and recurrent misunderstandings

The dominant noise terms vary by implementation but are recurrent. Reported sources include probe photodiode shot noise, electronic amplifier noise, laser intensity and frequency noise, thermal drifts, photodiode dark noise, AWG jitter, EOM $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$5 drift, atomic collision broadening, and transit-time broadening (Xie et al., 1 Jun 2026, Dixon et al., 2022). In low-frequency and low-data-rate receivers, cell charging, residual DC fields, and environmental magnetic fields are additional practical perturbations (Xie et al., 2024).

Bandwidth must be separated into several non-equivalent notions. Baseband bandwidth in AM/FM reception is approximately $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$6 (Anderson et al., 2018). Instantaneous optical-comb probe coverage can be approximately $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$7, with a time-resolution trade-off set by the comb spacing $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$8 and a full-scan time of $$E_{\rm AC}=\frac{\Delta f_2}{2\sqrt{2\,|\alpha|\,h\,\Delta f_1}}.$$9 (Dixon et al., 2022). In the sub-MHz Stark receiver, the effective detection bandwidth is approximately $\rho_{2,1}$0 because the coupling laser is scanned through resonance in $\rho_{2,1}$1 (Chen et al., 15 Jan 2025). In the Doppler-tracking architecture, the atomic instantaneous bandwidth is of order $\rho_{2,1}$2 or less, which is precisely why carrier offset tracking is required in high-mobility links (Xiang et al., 9 Jul 2026).

A recurrent misunderstanding is to equate calibration-free operation with the absence of front-end structures. The literature includes receivers described as not requiring antenna structures or providing optical circuit-free field pickup, but it also includes designs that deliberately use horn antennas, resonant cavities, and internal electrodes to increase coupling or realize low-frequency operation (Anderson et al., 2018, Xie et al., 1 Jun 2026, Xie et al., 2024). What remains calibration-free is the atomic conversion from spectroscopic observables to local field. The structures surrounding the vapor cell still determine how efficiently an external field is delivered to the atoms.

A second misunderstanding concerns “broadband” behavior. Rydberg receivers can address carriers from sub-MHz and $\rho_{2,1}$3 up to C-, K-, Ka-, Q-band and beyond by retuning the atomic transition or redesigning the cavity or antenna, but this does not imply arbitrarily wide instantaneous reception at fixed settings (Chen et al., 15 Jan 2025, Anderson et al., 2018, Xie et al., 1 Jun 2026). This distinction is central in both self-heterodyne comb readout and adaptive LO tracking.

6. Generalizations: multiband reception, polarimetry, and LO-free coherent recovery

The calibration-free concept generalizes in several orthogonal directions. One is frequency scaling. The S-band cavity work states that cavity and antenna designs can be scaled to other microwave bands such as X, Ku, and Ka by choosing resonant dimensions that scale as $\rho_{2,1}$4 and maintaining laser–microwave polarization alignment. The same source describes a compact package comprising a vapor cell, mini-cavity, micro-optics, diode lasers, photoreceiver, and FPGA DSP as a route to a field-strength meter or wireless-receiver front end that reports absolute electric field in real time (Xie et al., 1 Jun 2026).

A second direction is channel multiplicity. In the six-level hybrid receiver, four simultaneous RF channels are supported in one vapor-cell platform, exceeding the two channels of a parallel-only receiver and the three channels of a cascaded-only receiver. For a representative LO Rabi set $\rho_{2,1}$5 in units of $\rho_{2,1}$6, transmitted power $\rho_{2,1}$7, and $\rho_{2,1}$8, the reported ergodic sum-rate is approximately $\rho_{2,1}$9, corresponding to a $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$00 gain over CRS and a $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$01 gain over PRS (Shyamal et al., 13 Apr 2026).

A third direction is complete field-state recovery. In RF polarimetry with an optically wired atomic sensor, the RF state of polarization is mapped onto the Poincaré sphere through spectroscopic fingerprints. As the Stokes vector traverses a meridian, the atomic eigenenergy spectrum undergoes a continuous transformation, and the relative positions of eigenenergies are fixed by angular-momentum quantization. The reported framework is therefore universal and calibration free for systems with a single valence electron (Chilcott et al., 14 May 2026).

A fourth direction is coherent reception without an injected RF LO. In the DC-Stark-enabled receiver, a static bias mixes a near-degenerate Rydberg pair, closes a phase-sensitive loop, and yields an exact harmonic phase law in which the $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$02th probe harmonic carries the factor $$E_{\rm RF}=\frac{2\pi\hbar}{d_{\rm RF}}\Delta f_{\rm AT}.$$03. The signal phase is recovered from the harmonic phase, and the signal amplitude is recovered by inverting an injective harmonic response map. This establishes a minimal route to coherent reception of a single RF signal without an auxiliary RF LO in the atoms (Katkov et al., 31 Mar 2026).

Taken together, these results define the calibration-free Rydberg atomic receiver not as a single instrument, but as a family of atomic front ends in which the received field is mapped to an optical observable through known dipole moments, polarizabilities, and quantized level structure. The principal variations concern how the field is coupled into the atoms, how the optical signal is read out, and which communication observable—amplitude, phase, frequency, polarization, or channel index—is extracted.