Partner Frequencies in Constrained Systems

• Partner frequencies are defined as frequency components paired by governing constraints such as energy conservation and phase matching rather than chosen independently.
• Experimental validation involves techniques like joint spectral intensity mapping and coincidence counting to confirm symmetric pairings in resonator and SPDC systems.
• These constrained pairings enable advanced applications in quantum networking, multiplexed quantum secret sharing, and coordinated inter-agent communication.

“Partner frequencies” denotes frequency components that are linked by a governing constraint rather than chosen independently. In the photonic literature, the term most often refers to signal–idler frequencies generated together in a single spontaneous four-wave mixing or spontaneous parametric down-conversion event, with the pairing fixed by energy conservation and then filtered by cavity resonance, dispersion, and phase matching. In other literatures represented here, the same phrase or closely related usage denotes symmetric peaks in coupled-oscillator frequency distributions, a task-relevant interpersonal force band in dyadic haptic interaction, or common communication bands that two separated users must select under local jamming constraints (Yin et al., 2021, Henry et al., 2023, Mottola et al., 2019, Pietras et al., 2016, Colomer et al., 2022, Wehner, 22 Apr 2026).

1. Core definitions and pairing rules

Across the cited work, partner frequencies are defined by explicit pairing rules. In ring-resonator SFWM, the rule is symmetry about the pump; in broadband SPDC and DWDM discretization, it is symmetry about degeneracy; in nondegenerate cavity SPDC, it is a fixed signal–idler wavelength pair constrained by resonances; in coupled-oscillator theory, it is a symmetric pair of mean-frequency peaks; in haptic interaction, it is a partner-related frequency band rather than a signal–idler map (Yin et al., 2021, Henry et al., 2023, Mottola et al., 2019, Cabrejo-Ponce et al., 5 Aug 2025, Pietras et al., 2016, Colomer et al., 2022).

Context	Partner-frequency rule	Citation
SiN ring-cavity SFWM	$(+\mu)\leftrightarrow(-\mu)$ about the pump resonance	(Yin et al., 2021)
21 GHz silicon micro-resonator	$S_n=\omega_p/2\pi+n\,\mathrm{FSR}$ and $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$	(Henry et al., 2023)
Broadband SPDC with DWDM bins	$$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$	(Cabrejo-Ponce et al., 5 Aug 2025)
Monolithic cavity-enhanced SPDC	405 nm pump to 795 nm signal and 825 nm idler	(Mottola et al., 2019)
Symmetric Kuramoto subpopulations	frequency peaks at $+\varpi_0$ and $-\varpi_0$	(Pietras et al., 2016)
Dyadic haptic interaction	partner-related force band $[2.15,7]$ Hz	(Colomer et al., 2022)

The canonical photonic constraint is energy conservation. For resonant SFWM, the cited works write

$2\omega_p=\omega_s+\omega_i,$

and for SPDC they use

$\omega_p=\omega_s+\omega_i.$

When the spectrum is discretized around a pump or degeneracy frequency, this produces equal-and-opposite detunings: if one photon appears above the central frequency, its partner appears below it by the corresponding offset (Yin et al., 2021, Henry et al., 2023, Chen et al., 2021, Cabrejo-Ponce et al., 5 Aug 2025).

A recurring misconception is that energy conservation alone fixes the usable partner map. The cited cavity and fiber sources show that energy conservation determines only a candidate line or symmetry relation; the experimentally relevant partner frequencies are the subset that also satisfy resonance, dispersion, and phase-matching conditions (Yin et al., 2021, Mottola et al., 2019, Torre-Robles et al., 2021).

2. Resonator-based partner frequencies in integrated SFWM

In the SiN ring-cavity source, the pump mode is labeled $\mu=0$, and the ideal partner rule is

$S_n=\omega_p/2\pi+n\,\mathrm{FSR}$0

so the signal in mode $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$1 is paired with the idler in mode $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$2. The resonant frequencies are expanded as

$S_n=\omega_p/2\pi+n\,\mathrm{FSR}$3

with integrated dispersion

$S_n=\omega_p/2\pi+n\,\mathrm{FSR}$4

The device was engineered for low anomalous dispersion; the measured FSR near 1550 nm was about 145 GHz, the extracted coefficients were $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$5 and $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$6, and the frequency mismatch across tens of modes remained below the cavity linewidth of about 132 MHz. Frequency-resolved coincidence counting over a full $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$7 map confirmed 42 correlated partner-frequency pairs, including 37 continuous mode pairs, corresponding to a correlated bandwidth of 51.25 nm in the signal band (Yin et al., 2021).

The phase-matching condition in that device was written as

$S_n=\omega_p/2\pi+n\,\mathrm{FSR}$8

This makes the resonator-specific partner rule precise: symmetric mode numbers remain good partners only when the sum of the integrated-dispersion terms, together with pump-induced self-phase modulation, keeps the mismatch small. The measured joint spectral intensity showed dominant coincidences along the diagonal corresponding to the symmetric pairings $S_n=\omega_p/2\pi+n\,\mathrm{FSR}$9, while off-diagonal terms were much weaker and interpreted as non-phase-matched background or residual accidentals (Yin et al., 2021).

The 21 GHz silicon micro-resonator implements the same symmetry in a denser comb. Its resonances are labeled

$I_n=\omega_p/2\pi-n\,\mathrm{FSR}$0

with measured $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$1 GHz, resonance linewidth about 600 MHz FWHM, and $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$2. The source provides more than 70 frequency channels separated by 21 GHz, and correlated pairs were observed from $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$3 to $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$4. The same narrow spacing enabled parallel manipulation: 34 single-qubit gates were applied in parallel to 17 frequency-bin maximally entangled qubit pairs, with crosstalk below $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$5 when two guard modes were left between consecutive qubits (Henry et al., 2023).

These resonator studies jointly establish a characteristic integrated-photonics meaning of partner frequencies: resonant bins symmetrically disposed around the pump, with the nominal pairing rule set by energy conservation and the observed pairing range set by linewidth-scale tolerance to dispersion-induced mismatch (Yin et al., 2021, Henry et al., 2023).

3. Nondegenerate SPDC, bin discretization, and hyperentangled partner structures

In the monolithic ppKTP optical parametric oscillator designed for the $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$6Rb D1 line, partner frequencies are explicitly nondegenerate. The crystal is engineered for type-II conversion of 405 nm pump photons to 795 nm signal photons and 825 nm idler photons. The source therefore does not generate two photons at 795 nm; instead, one photon is tuned to the $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$7 D1 region near 794.979 nm and the other is its longer-wavelength partner near 825 nm. The allowed pairs are further restricted by phase matching, signal-cavity resonance, idler-cavity resonance, and triple resonance of the pump. Because signal and idler have different free spectral ranges, the spectrum is clustered rather than uniformly paired. Around the D1 line, the signal cavity transmission has FSR = 16 GHz, the central emission peak contributes 75% of the overall cluster intensity, and the linewidth of that dominant emission line is $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$8 MHz FWHM. Fine tuning by mechanical strain shifts the selected pair by more than 2 GHz, while passive drift is about 10 MHz per hour (Mottola et al., 2019).

Broadband telecom SPDC sources typically use symmetry about degeneracy rather than a fixed nondegenerate wavelength pair. In the all-fiber Sagnac source for multiplexed quantum secret sharing, the generated state is centered at the degeneracy wavelength $I_n=\omega_p/2\pi-n\,\mathrm{FSR}$9 nm, equivalently $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$0 THz. After discretization, partner channels are written as

$$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$1

so each channel pair is symmetric with respect to the degeneracy frequency. On the 200 GHz ITU grid used in the experiment, the 70 nm source bandwidth yields 22 partner channel pairs; with 100 GHz DWDM, the same source would provide at least 44 pairs of frequency channels. The proof-of-principle QSS demonstration used channels 13 and 60, corresponding to $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$2, and reported fidelities of 93.0%, 92.9%, 90.3%, and 95.4% for the four encoded states, with average QBER 5.3% and key fraction 0.4 bits/photon (Cabrejo-Ponce et al., 5 Aug 2025).

The periodically poled silica fiber hyperentanglement source uses a cw 778 nm pump and degenerate downconversion at 1556 nm. Here the selected partner bins are centered symmetrically around $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$3 at detunings $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$4 THz, $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$5 THz, $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$6 THz, and $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$7 THz, each with 0.4 THz top-hat passband width. The operational frequency-bin Bell-like factor is

$$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$8

which expresses coherent superposition of the two swapped assignments of the upper and lower partner bins to the two output modes. The inferred frequency-subspace concurrence is approximately 0.988, the polarization concurrence is $$|j,j\rangle_{s,i}=|\omega_0+j\Delta\omega,\omega_0-j\Delta\omega\rangle$$9, and the global-state fidelity is bounded by $+\varpi_0$0 (Chen et al., 2021).

A distinct hyperentangled construction maps pulse-mode entanglement into discrete frequency-bin pairing. Its final frequency-bin factor is

$+\varpi_0$1

so the partner relation is not a single deterministic assignment but a coherent superposition of the two bin orderings. The measured joint spectral intensity displays two distinct pairs of lobes; the fitted frequency-bin separation is $+\varpi_0$2 THz and the bin width is $+\varpi_0$3 THz, corresponding to approximately 11 nm separation and 3 nm width around 1550 nm (Chiriano et al., 2023).

These sources demonstrate that partner frequencies need not mean a single architectural pattern. Depending on the platform, the partner relation may be symmetric about a pump, symmetric about degeneracy, clustered by nondegenerate cavity resonances, or embedded in a higher-order hyperentangled structure (Mottola et al., 2019, Chen et al., 2021, Chiriano et al., 2023, Cabrejo-Ponce et al., 5 Aug 2025).

4. Identification, modeling, and experimental constraints

The principal experimental identifier of partner frequencies in photon-pair sources is the joint spectral intensity or an equivalent spectrally resolved coincidence measurement. In the SiN ring-cavity experiment, the pump was suppressed with a fiber Bragg grating, the output was split into signal and idler arms, and each arm was filtered with a tunable bandpass filter of 0.12 nm passband. Coincidences were recorded with superconducting nanowire single-photon detectors and a time-to-digital converter using a coincidence window $+\varpi_0$4 ns and 10 s integration per mode pair. To suppress uncorrelated background, the accidental coincidence count

$+\varpi_0$5

was subtracted. The resulting $+\varpi_0$6 coincidence map provided the direct experimental partner-frequency map: each bright diagonal element corresponded to one mode pair $+\varpi_0$7 (Yin et al., 2021).

In the birefringent photonic crystal fiber source, partner frequencies are identified by combining energy-conservation mapping, polarization selection, and spectrally resolved coincidences. The exact energy relation is

$+\varpi_0$8

and the authors use the transformation $+\varpi_0$9 to overlay measured signal and idler spectra. The source supports up to six distinct SFWM processes, labeled $-\varpi_0$0 through $-\varpi_0$1, each with its own polarization assignment and phase-matching contour. Because phasematching is polarization-dependent, the six processes are generally spectrally distinct, and in some cases, particularly process $-\varpi_0$2, two signal-mode peaks and two idler-mode peaks may appear for a single pump frequency (Torre-Robles et al., 2021).

In the PPSF hyperentanglement experiment, full frequency-domain tomography was not used; instead, frequency entanglement was inferred from Hong–Ou–Mandel interference. For top-hat bins of width $-\varpi_0$3 centered at detuning $-\varpi_0$4, the coincidence probability was modeled as

$-\varpi_0$5

For channel 1, the fit used $-\varpi_0$6 THz and $-\varpi_0$7 THz; measured visibilities exceeded 96.9%, with average 98.9%, and the anti-bunching configuration at $-\varpi_0$8 served as the operational witness of antisymmetric frequency-bin entanglement (Chen et al., 2021).

The monolithic OPO source used seeded cavity transmission and stimulated DFG rather than a full JSI. Around the relevant cluster, the DFG scan showed a strong central peak plus two weaker doublets each separated by one FSR, and no measurable DFG was observed outside the displayed window. The detected idler spectrum was the product of the 226(1) MHz Lorentzian source line and the 274(4) MHz Lorentzian idler filter cavity line, showing that stringent idler filtering effectively projected the experiment onto one chosen partner-frequency channel (Mottola et al., 2019).

These studies converge on a common methodological point: partner frequencies are experimentally real only after mode selectivity is imposed. Energy conservation defines a locus; phase matching, cavity resonances, filter resolution, pump rejection, and accidental-background control determine which elements of that locus survive as identifiable, high-contrast partners (Yin et al., 2021, Mottola et al., 2019, Torre-Robles et al., 2021, Chen et al., 2021).

5. Parallel processing, networking, and coordinated band selection

Once a source provides many partner-frequency channels, the pairing rule becomes an address space for quantum information processing. In the 21 GHz silicon micro-resonator, adjacent partner pairs $-\varpi_0$9 and $[2.15,7]$0 were grouped into frequency-bin Bell-like qubits, and an electro-optic frequency-domain gate in an $[2.15,7]$1 configuration was driven at $[2.15,7]$2 to couple neighboring resonator lines. This enabled parallel and arbitrary qubit transformations, full quantum state tomography on 17 frequency-entangled qubit pairs, fidelity $[2.15,7]$3 for the example $[2.15,7]$4, fidelity above 0.8 for 14 pairs, and QBER below the 11% secure-key threshold for 12 of the 17 accessible pairs. The same partner-channel structure supported a node-free fully connected 5-user network using 10 entangled photon pairs, matching the $[2.15,7]$5 user-user links (Henry et al., 2023).

The frequency-subspace QSS protocol extends the same logic to multiplexed secret sharing. Each partner-frequency channel pair $[2.15,7]$6 is treated as an independent subspace, and the dealer applies a frequency-dependent phase $[2.15,7]$7 to the idler branch so that each partner pair hosts an independent QSS session. The source bandwidth of about 70 nm supports 22 partner channel pairs on the 200 GHz grid used experimentally, and the abstract states that the method could be extended to more than 40 frequency bins with adequate dense-wavelength division multiplexed filters. One partner pair, channels 16 and 57, was consumed for stabilization, while a different partner pair, channels 13 and 60, carried the demonstrated QSS states (Cabrejo-Ponce et al., 5 Aug 2025).

A different operational meaning appears in the coordinated frequency-selection game against distributed jammers. There are $[2.15,7]$8 bands, each party sees a safe set of size $[2.15,7]$9, and success requires that Alice and Bob output the same band in the intersection of their local safe sets. The paper solves the optimal classical strategy exactly: the optimum deterministic rule is the greedy strategy $2\omega_p=\omega_s+\omega_i,$0, and the corresponding success probability is

$2\omega_p=\omega_s+\omega_i,$1

With pre-shared entanglement, the parties can coordinate better. For $2\omega_p=\omega_s+\omega_i,$2, one shared Bell pair already outperforms the classical optimum for all spectrum sizes $2\omega_p=\omega_s+\omega_i,$3, and the asymptotic gain is

$2\omega_p=\omega_s+\omega_i,$4

that is, a 5.4% advantage asymptotically in $2\omega_p=\omega_s+\omega_i,$5 (Wehner, 22 Apr 2026).

Taken together, these papers show that partner frequencies can function either as pre-existing spectral correlations exploited for multiplexing or as coordination targets that two parties must select jointly under local constraints. In both cases, the decisive resource is structured correlation across contexts rather than isolated per-channel performance (Henry et al., 2023, Cabrejo-Ponce et al., 5 Aug 2025, Wehner, 22 Apr 2026).

6. Related usages in oscillator theory and haptic interaction

In coupled-oscillator theory, the phrase is not used explicitly, but the paper’s central concept is a pair of symmetric frequency peaks at $2\omega_p=\omega_s+\omega_i,$6 and $2\omega_p=\omega_s+\omega_i,$7. Two symmetric Kuramoto subpopulations with Lorentzian distributions

$2\omega_p=\omega_s+\omega_i,$8

reduce under the Ott–Antonsen ansatz to the same dynamics as a single population with a bimodal frequency distribution when $2\omega_p=\omega_s+\omega_i,$9. The reduced equations

$\omega_p=\omega_s+\omega_i.$0

show that the peak separation and the relative phase between the two subpopulation order parameters govern the transition among incoherence, partially synchronized fixed points, and stable limit cycles. The paper’s conclusion is that stability, dynamics, and bifurcations of a symmetric two-population system are equivalent to a single population with bimodal frequency distribution, but that this equivalence does not readily generalize to multimodal or multiple-subpopulation systems (Pietras et al., 2016).

In dyadic haptic interaction, partner frequencies refer to a frequency band in the interaction forces of two physically coupled people. Frequency-domain Granger–Geweke causality applied to estimated individual force outputs identified a band $\omega_p=\omega_s+\omega_i.$1 Hz in which inter-partner causal influence was asymmetric when information exchange was indispensable, absent when such exchange was unnecessary, and correlated with better task performance. The overt movement frequency in the task was below 1 Hz, so the partner-related band was distinct from the movement rhythm. Performance correlations were strongest in the sub-bands $\omega_p=\omega_s+\omega_i.$2 Hz and $\omega_p=\omega_s+\omega_i.$3 Hz, while no significant correlations were found in $\omega_p=\omega_s+\omega_i.$4 Hz or $\omega_p=\omega_s+\omega_i.$5 Hz (Colomer et al., 2022).

These usages delimit the concept. In oscillator theory, partner frequencies are paired peaks in a frequency distribution; in haptics, they are partner-related bands mediating interaction; in photonics, they are spectrally correlated signal–idler channels. The common structure is coupling across agents or modes, but the mechanism differs: symmetry and bifurcation in one case, predictive force exchange in another, nonlinear optical creation and filtering in the third (Pietras et al., 2016, Colomer et al., 2022).

A final conceptual boundary follows from the photonic examples. Symmetry in index or detuning does not imply uniform pair strength. In the SiN ring cavity, correlations remain symmetric about the pump in mode number, but dispersion and filtering determine which symmetric pairs are actually accessible or strong enough to observe; coincidence counts decline with increasing $\omega_p=\omega_s+\omega_i.$6, with unusually low counts around the 20th mode because mismatch increases there (Yin et al., 2021). Likewise, in the monolithic OPO, the existence of a 405 nm $\omega_p=\omega_s+\omega_i.$7 795 nm + 825 nm energy relation does not mean that all compatible pairs are emitted equally; the dominant central cluster mode contributes 75% of the cluster intensity because the signal and idler resonances are best aligned there (Mottola et al., 2019).

Partner frequencies, therefore, is best understood as a family of constrained pairings whose exact form depends on the governing dynamical law—energy conservation and phase matching in nonlinear optics, symmetry and coupling in oscillator ensembles, or interaction-specific spectral causality in dyadic force exchange.