The coupling coefficient sensing characteristics of the CP-based architecture.

A collaborative effort between Peking University's School of Electronics and Xi'an Jiaotong University's (XJTU) School of Electrical Engineering recently led to a breakthrough in non-Hermitian wireless sensing, which has been published in the internationally renowned journal Science Advances under the title Critical point-based wireless sensors enabling tiny perturbation detection.

The research team proposed the concept of a universal non-Hermitian critical point (CP) and designed a novel electronic circuit architecture for wireless sensing. This innovation achieves high-precision, long-distance, and high-signal-to-noise ratio (SNR) real-time wireless monitoring of weak signals.

Wireless sensing technology is free from the constraints of physical wire connections, giving it immense potential across various fields. These include power grid monitoring in enclosed or complex environments, high-end aerospace equipment, advanced implantable healthcare devices, and embodied AI/internet of things sensor networks.

However, conventional wireless sensing architectures are limited by inherent loss characteristics, making long-distance, high-precision detection of minute signals incredibly difficult. While sensing architectures based on parity-time (PT) symmetry can boost performance, they require a strict balance of gain and loss configurations. This imposes harsh requirements on component parameters and strong coupling coefficients. Furthermore, thermal noise amplifies drastically near the exceptional point (EP), severely limiting practical applications.

To overcome these limitations, the research team proposed a universal non-Hermitian critical-point wireless sensing architecture that utilizes an unbalanced gain-loss configuration.

Compared to traditional and PT-symmetric architectures, this new framework operates under more relaxed constraints and exhibits unique, rich dynamic characteristics. It not only achieves an ultra-high quality factor (Q-factor) and extended wireless sensing distances, but also solves the long-standing trade-offs between sensing sensitivity, SNR, and coupling strength.

Experimental data demonstrates the extraordinary capabilities of this new architecture. It can resolve changes in the coupling coefficient as low as 1.92×10⁻⁴, improving the detection limit by more than sevenfold compared to PT-symmetric architectures. Additionally, because the coupling coefficient is independent of the characteristic frequency, it eliminates frequency-response interference caused by changes in capacitance.

When used to detect tiny asymmetric capacitance variations, it can sense perturbations as low as 2.5×10⁻⁵ (corresponding to a capacitance change of approximately 0.6 femtofarads). This marks an improvement of two to three orders of magnitude over previously reported results.

Even in real-world environments with complex vibration noise, this sensing architecture reliably captures minute asymmetric capacitance changes for long-distance, real-time signal monitoring.

Furthermore, the architecture can be extended to fields such as wireless power transfer (WPT), offering a new pathway for high-efficiency wireless power supply. This research transitions non-Hermitian dynamics theory into practical wireless sensing applications, providing a fresh theoretical framework and technical solution for developing next-generation wireless electronic sensor devices and systems.