The Role of Exceptional Points and Transmission Peak Degeneracies in Non-Hermitian Sensing

Technical Summary: The Role of Exceptional Points and Transmission Peak Degeneracies in Non-Hermitian Sensing

Problem Statement
Non-Hermitian exceptional points (EPs) have been proposed as a mechanism for enhanced sensing due to their sublinear eigenvalue splitting ($\Delta\lambda \propto \epsilon^{1/n}$) near a perturbation $\epsilon$. However, practical implementation of EP-based sensors faces two fundamental hurdles. First, EPs are isolated degeneracies in high-dimensional parameter spaces; any drift in nuisance (non-sensing) parameters instantly lifts the degeneracy, destroying the square-root splitting and the associated enhanced response. Second, the eigenbasis collapse at EPs inherently amplifies noise (quantified by the divergence of the Petermann factor), which fundamentally limits the signal-to-noise ratio (SNR).

While recent work suggested that transmission peak degeneracies (TPDs)—degeneracies in the transmission spectrum rather than the eigenspectrum—could offer square-root splitting without eigenbasis collapse, the field lacks a unified theoretical framework. Existing treatments of TPDs are fragmented, lacking systematic figures of merit, design principles for robust operation, or a clear mapping of their relationship to EPs across different symmetry classes.

Methodology
The authors develop a comprehensive semiclassical theory for two-dimensional TPDs and validate it experimentally using a tunable cavity-magnonics platform.

• Theoretical Framework: The system is modeled as a magnon-photon dimer governed by a non-Hermitian dynamical matrix $\tilde{A}$. The authors derive analytic expressions for the eigenspectrum, transmission spectrum, and the conditions for TPDs. They define TPDs as points where the discriminant of the transmission extrema equation vanishes ($Disc=0$) simultaneously with the condition $\tilde{q}=0$, distinguishing them from simple transmission extrema degeneracies (TEDs).
• Experimental Platform: The experiment utilizes a hybrid system consisting of a 3D microwave cavity (photon mode) and a yttrium iron garnet (YIG) sphere (magnon mode). The modes are coupled via a loop containing tunable amplifiers and a digital phase shifter, creating an effective synthetic gauge field with a controllable coupling phase $\phi$.
• Control and Validation: The platform allows for in situ control of mode frequencies ($f_c, f_y$), dissipation rates ($\kappa_c, \kappa_y$), and complex coupling ($J, \phi$). The authors systematically explore six representative configurations spanning $PT$-symmetric ($\phi=0$), anti-$PT$-symmetric ($\phi=\pi$), and anyonic-$PT$-symmetric ($\phi=\pi/2$) regimes. They map the parameter space to locate EPs and TPDs and validate the square-root splitting of transmission peaks along specific trajectories ($\tilde{q}=0$) that intersect both EPs and TPDs.

Key Contributions and Results

1. Unified Theoretical Framework: The paper establishes a unified model connecting EPs and TPDs. It demonstrates that while EPs are stationary in parameter space for a fixed phase $\phi$, TPDs are tunable and can be moved by adjusting the average dissipation rate $\tilde{\kappa}_c$. The authors provide analytic figures of merit for sensor design, including the scaling coefficient for peak splitting, the Petermann factor (noise amplification), and thermal noise efficiency.

2. Experimental Validation of TPDs: The authors experimentally confirm that TPDs exhibit square-root frequency splitting ($\Delta\nu \propto \sqrt{\epsilon}$) similar to EPs. Crucially, they show that TPDs maintain a complete eigenbasis, resulting in a finite Petermann factor and improved SNR compared to EPs. The experiments cover three distinct symmetry regimes, confirming the theoretical predictions for peak locations and splitting behavior.

3. Robustness Against Nuisance Drift: A critical finding is the analysis of TPD vulnerability to nuisance parameter fluctuations. While TPDs generally retain square-root splitting even when the sensing path is displaced by noise (unlike EPs where degeneracy is lifted), most TPDs suffer from a "cube-root" scaling response to nuisance parameters ($\Delta\nu \propto \delta^{1/3}$), which can dominate the signal.

   • Robust TPDs: The authors identify a specific configuration (a "robust TPD") where the cube-root nuisance scaling coefficient vanishes. For the $PT$-symmetric case ($\phi=0$), this occurs at $\tilde{\kappa}_c = 2$. In this configuration, the nuisance response becomes linear, and the transmission peak splitting remains suppressed against fluctuations.
   • Third-Order Degeneracy: The paper highlights that the third-order degeneracy of all TPDs provides a fundamental advantage: even if nuisance fluctuations displace the sensing path from the exact TPD, the path still crosses a second-order Transmission Extrema Degeneracy (TED), preserving the square-root splitting response.

4. Design Principles: The authors derive analytic figures of merit to guide the selection of optimal operating points. These include the distance to the nearest instability transition, the distance to the nearest EP (which dictates the Petermann factor), and the specific dissipation rates required to achieve robust TPDs.

Significance
The paper claims to establish a foundational theoretical and experimental framework for TPD-based non-Hermitian sensing. By clarifying the relationship between EPs and TPDs and providing a unified parameter landscape, the work addresses the fragmented nature of previous TPD studies. The identification of "robust TPD" configurations offers a concrete design principle for mitigating the notorious fragility of degeneracy-based sensors to nuisance fluctuations. The authors posit that their platform and formalism serve as a versatile testbed for exploring non-Hermitian dynamics and provide the necessary design rules for implementing practical, high-performance sensors that leverage TPDs while avoiding the noise amplification and instability issues associated with EPs.