Non-Hermitian sensing via end-to-end Green's functions
AI Breakdown
Get a structured breakdown of this paper — what it's about, the core idea, and key takeaways for the field.
Abstract
Sensing hinges on two key attributes: high sensitivity and resilience to noise. Here, we present a scheme that unites these features within a single framework. By harnessing the exponential amplification of end-to-end Green's functions-a direct signature of the non-Hermitian skin effect-our scheme achieves an exponential sensitivity scaling of exp(αL) with α a model-specific constant and L the system size. Notably, this enhanced sensitivity is protected by spectral winding numbers and endures under disorder when the non-Hermitian topological phase remains intact. We also explore its feasibility in synthetic non-Hermitian platforms, providing a pathway toward highly sensitive and noise-tolerant sensing.