Demodulation of Fabry-Perot sensors using random speckles

Area of Science:

• Optical sensing
• Fiber optics
• Signal processing

Background:

• Fabry-Perot (FP) sensors are widely used for various measurements.
• Traditional demodulation methods for FP sensors face challenges like fringe-order ambiguity and limited dynamic range.
• Developing advanced demodulation techniques is crucial for improving FP sensor performance.

Purpose of the Study:

• To propose and demonstrate a novel demodulation scheme for FP sensors using random speckles.
• To investigate the potential of speckle pattern analysis for high-resolution and wide-dynamic-range measurements.
• To address the fringe-order ambiguity problem in FP sensor interrogation.

Main Methods:

• Utilizing a multimode fiber to interrogate the FP transmission spectrum.
• Analyzing the significant variations in generated speckle patterns corresponding to tiny spectral changes.
• Employing a convolutional neural network (CNN) based demodulation algorithm.

Main Results:

• Achieved a pressure resolution of 0.001 MPa with an open-cavity FP sensor.
• Demonstrated precise distinction of spectral differences in neighboring orders due to high speckle sensitivity.
• Successfully solved the fringe-order ambiguity problem, significantly improving the dynamic measurement range.

Conclusions:

• Random speckles offer a new, effective method for demodulating FP sensors.
• The speckle-based scheme provides a balanced approach to resolution, dynamic range, speed, and cost.
• This technique opens new possibilities for advanced optical sensing applications.