Export citation

Export citation

Choose format for download:

Download Citation

    Optimizing the sensitivity-noise trade-off in non-Hermitian sensing via off-exceptional-deficiency operation

    Shangxuan Li and Bin Guo*

    • *Contact author: binguo@whut.edu.cn

    Phys. Rev. B 114, 115126 – Published 21 August, 2026

    DOI: https://doi.org/10.1103/w3ph-rl5m

    Abstract

    A central challenge in non-Hermitian sensing is that spectral singularities simultaneously amplify both the signal and environmental noise. We address this predicament in a double-chain Hatano-Nelson model featuring unidirectional interlayer coupling. At the exceptional-deficiency (ED) limit, the system exhibits a macroscopically degenerate complex spectrum and a pronounced non-Hermitian skin effect (NHSE), yielding a sensitivity that scales exponentially with lattice size N while remaining robust across a six-order-of-magnitude detuning range. By introducing diagonal spatial disorder, we demonstrate that the NHSE is progressively suppressed, whith eigenspace cosine similarity analysis quantifying a well-defined fault-tolerance threshold. To reconcile the sensitivity-noise trade-off, we delineate at-ED and off-ED operating regimes. While the at-ED configuration imposes fractional-order noise amplification [signal-to-noise ratio (SNR) ∝δ−1/2] that saturates at a suboptimal plateau, migrating to the off-ED regime eliminates this geometric singularity and restores a linear scaling law (SNR ∝δ−1), achieving an SNR enhancement of several orders of magnitude. Crucially, this improvement is achieved while fully preserving the exponential sensitivity scaling, albeit at a slightly reduced absolute sensitivity compared with the strict at-ED limit. Our findings establish the off-ED framework as a concrete paradigm for next-generation topological sensors that reconcile extreme sensitivity with robust noise immunity.

    Physics Subject Headings (PhySH)

    Authorization Required

    We need you to provide your credentials before accessing this content.

    References (Subscription Required)

    Outline

    Information

    Sign In to Your Journals Account

    Filter

    Filter

    Article Lookup

    Enter a citation