High-bandwidth processing of heterodyne spectroscopic signals for remote and extractive chemical sensing

Abstract

This works explores the noise sources and real-time processing of high-bandwidth signals of heterodyne detection based laser spectroscopic techniques for in-field trace gas sensing applications.

Two spectroscopic techniques are studied in detail: chirped laser dispersion spectroscopy (CLaDS) and dual comb spectroscopy (DCS). Chapter 1 is an introduction to absorption and dispersion spectroscopy techniques suitable for trace gas sensing applications, represented by direct absorption spectroscopy (DAS), CLaDS and DCS. Chapter 2 discusses the operating principles of CLaDS and its variations including simultaneous ranging and heterodyne enhancement, followed by Chapter 3 which quantifies the detection limits of the CLaDS techniques under white noise. Chapter 4 focuses particularly on speckle formation caused by scattering in differential measurements using heterodyne detection, commonly observed in standoff detection applications. The modeling and experimental verification of speckle noise were performed using CLaDS as the case study example, but the results can be generalized to other differential phase-sensitive signals. Chapter 5 includes system implementation of a CLaDS sensor for natural gas leak detection. Design concerns and implementation methods are discussed, including real-time processing of the CLaDS signal on a FPGA platform. The results from characterization tests and field deployment are examined and are related to the noise models established in previous chapters. Chapter 6 centers around signal processing of DCS signals corrupted by frequency noise. Coherent averaging and frequency correction methods are introduced. Experimental results from in-lab characterization and field campaigns are discussed to demonstrate the noise mitigation and its physical limitations using high bandwidth real-time processing.