We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.SP

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Signal Processing

Title: AI Algorithm for Mode Classification of PCF SPR Sensor Design

Abstract: Photonic Crystal Fiber design based on surface plasmon resonance phenomenon (PCF SPR) is optimized before it is fabricated for a particular application. An artificial intelligence algorithm is evaluated here to increase the ease of the simulation process for common users. COMSOL MultiPhysics is used. The algorithm suggests best among eight standard machine learning and one deep learning model to automatically select the desired mode, chosen visually by the experts otherwise. Total seven performance indices: namely Precision, Recall, Accuracy, F1-Score, Specificity, Matthew correlation coefficient, are utilized to make the optimal decision. Robustness towards variations in sensor geometry design is also considered as an optimal parameter. Several PCF-SPR based Photonic sensor designs are tested, and a large range optimal (based on phase matching) design is proposed. For this design algorithm has selected Support Vector Machine (SVM) as the best option with an accuracy of 96%, F1-Score is 95.83%, and MCC of 92.30%. The average sensitivity of the proposed sensor design with respect to change in refractive index (1.37-1.41) is 5500 nm/RIU. Resolution is 2.0498x10^(-5) RIU^(-1). The algorithm can be integrated into commercial software as an add-on or as a module in academic codes. The proposed novel step has saved approximately 75 minutes in the overall design process. The present work is equally applicable for mode selection of sensor other than PCF-SPR sensing geometries.
Comments: 22 pages, 11 figures
Subjects: Signal Processing (eess.SP)
Cite as: arXiv:2107.06184 [eess.SP]
  (or arXiv:2107.06184v2 [eess.SP] for this version)

Submission history

From: Mayank Goswami [view email]
[v1] Fri, 9 Jul 2021 04:59:13 GMT (2215kb)
[v2] Mon, 19 Jul 2021 05:30:02 GMT (2215kb)

Link back to: arXiv, form interface, contact.