Robust Computational Spectral Sensing via Orthogonal Frequency Encoding
ABSTRACT High‐fidelity spectroscopy in real‐world environments is limited by ambient interference and the inherent trade‐off between acquisition speed and signal‐to‐noise ratio (SNR). Conventional time‐division multiplexing approaches often fail as they lack intrinsic mechanisms to decouple target signals from background illumination. Here, we propose orthogonal frequency encoding, a robust framework that shifts spectral acquisition from serial scanning to a parallelized frequency‐domain architecture. By encoding each spectral channel with a distinct orthogonal frequency kernel, it allocates carriers to frequency gaps to evade structured environmental noise. As proof of concept, we use a programmable LED array to demonstrate scan‐free computational spectral reconstruction based on this framework. Under the tested SMPS interference, the system achieves a reconstruction RMSE of approximately 0.0261 at a modulation SNR of −11.77 dB using a 1000 μs acquisition window. This work establishes a fast, low‐cost strategy for spectral sensing in challenging environments.
Authors
- Yijia Zeng
- Sailing He (ORCID: https://orcid.org/0000-0002-3401-1125)
- Tingbiao Guo (ORCID: https://orcid.org/0000-0002-4036-164X)
- Feng Zhou (ORCID: https://orcid.org/0000-0001-8182-595X)
- Quanzhou Long
- Yafan Jin
- Zijian Lin
- Zhi Zhang
Institutions
- Zhejiang University (CN)
- KTH Royal Institute of Technology (SE)
Publication Details
- Journal
- Nanophotonics
- Published
- 2026-11-17
- DOI
- https://doi.org/10.1002/nap2.70370
- Primary Topic
- Spectroscopy and Laser Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00