Fiber optic networks capable of supporting a large pool of subscribers, many simultaneous users, and high data rates are receiving heightened interest as solutions to a growing communications need. The experiments reported in this study constitute the first experimental demonstration of a novel bipolar equivalent code-division multiple-access (CDMA) scheme. The sophisticated encoding increases noise tolerance, provides user security, and enables network flexibility. The scheme is based on an established bipolar radio frequency (RF) technique adapted to the unipolar optical domain. Whereas the phase of an RF signal can be readily detected, the high carrier frequency ( ∼ 200 THz at 1.5 mum) of an optical wave necessitates that optical signals be detected and processed solely by intensity. Asynchronous operation makes the CDMA scheme data rate independent, while all-optical implementation avoids the bandwidth limitations imposed by electrical processing. A proof-of-principle experiment was conducted by spectrally encoding an erbium-doped superfluorescent fiber source (SFS) using a diffraction grating and an amplitude mask. The optical properties of the system were measured and the bipolar correlation of codes was verified. The practical implementation of the scheme was investigated by the design, construction, and operation of a fiber-based testbed. Correlation measurements performed with modulated signals confirmed that the scheme can recover a binary information symbol while rejecting multiple access interference. A theoretical analysis of the optical correlation process was conducted, which identified key optical parameters important to future implementations. The theory of excess noise associated with the photodetection of a thermal source was considered, followed by noise measurements of a light bulb and the erbium-doped SFS used for spectral encoding. Finally, the ability of the proposed scheme to effectively transmit data was investigated. Signal-to-noise-ratio and eye diagram measurements of the testbed were recorded. Bit-error rate (BER) measurements were conducted for various network scenarios, including operation with one and two active users, as well as with varying amounts of broadband optical interference. The analysis of the BER measurements addresses the limitations due to excess noise, the impact of imperfect correlation performance) and the anticipated capacity for multiple users.

We constructed a 3.0 meter diameter f/1.5 Liquid Mirror Telescope (LMT) between 1990 and 1994 at the NASA Johnson Space Center, Houston, Texas. We have subsequently operated it since 1995 at the NASA Orbital Debris Observatory (NODO), Cloudcroft, NM. Employing an inexpensive rotating container of mercury as its primary parabolic mirror, the NASA LMT is a cost-effective alternative to telescopes utilizing glass mirrors. We detail criteria for mirror construction including environmental considerations via Hg vapor emission analysis. We describe performance optimization to the NODO site seeing limit of 0.8 arcseconds FWHM via analysis of perturbations to image quality from mirror angular velocity stability, dynamic balance, rotational axis tilt, and prime focus lateral and tilt displacements. We detail the behavior of the two prominent mirror surface wave phenomena---spiral and concentric forms. We demonstrate that the former probably results from vorticity in the air boundary layer above the mirror and show diffraction effects from the latter. We describe mirror stabilization in terms of boundary layer theory. The prime focus NASA-LMT utilizes corrective optics yielding a field of 46 arcminute diameter. Utilizing Micro-Channel-Plate (MCP) intensified video cameras we have obtained 750 hours of zenith staring orbital object event data with a limiting object diameter of approximately 1 cm at 1000 km altitude and 0.1 albedo. We have extended to 17.75 the lower magnitude limit of optical detections among the telescopes employed for orbital object surveys, further demonstrated the incompleteness of the SATCAT, and corroborated results of RADAR employed in orbital object detection. Utilizing CCDs we have conducted a 135 night broadband and multi-narrowband survey of 20 square degrees of sky at high galactic latitude down to a limiting magnitude of ∼22.0. The survey data will yield information on object morphology, spectral classifications, and large-scale structure to a redshift (z) of 0.5 with an accuracy of Deltaz ≤ 0.02. Broadband images from this survey are presented, demonstrating that the NASA-LMT optical performance is comparable to conventional telescopes of equivalent size located at a similar site.

This thesis describes the development of digital signal processor (DSP) and quantum cascade laser (QCL) technology for compact, standalone, integrated, trace gas sensors capable of gas sensing at part-per-billion (ppb) concentration levels. Such sensitivity levels provide the ability to perform precise environmental and emissions monitoring, medical biomarker analysis, and biological/toxic agent detection. Current sensors based on tunable diode laser absorption spectroscopy (TDLAS) frequently require complex optical configurations or cryogenics to produce the proper mid-infrared optical frequencies, a full personal computer based data acquisition system to process data, and assorted nonintegrated support electronics for complete control of the system. In contrast, the developed system employs pulsed distributed feedback (DFB) QCLs to provide frequencies in the mid-infrared wavelengths targeting specific molecular ro-vibrational lines for simple direct absorption spectroscopy, while a custom DSP implementation provides integrated control and processing functions without sacrificing performance. Results show comparable sensitivities compared to traditional techniques, while simultaneously producing compact, robust, low power solutions.

Terahertz time-domain spectroscopy (THz-TDS) offers a range of unique imaging modalities due to the broad bandwidth, sub-picosecond duration, and phase-sensitive detection of the THz pulses. Previous research in the THz field primarily focused on improving the optoelectronics and on qualitative investigation of suitable applications. In this thesis, we use a quantitative approach to explore and extend the boundaries of the system. First, the possibility exists to combine spectroscopic characterization and/or identification with imaging because the THz radiation is broadband in nature. We describe a robust algorithm for extracting the optical constants and thickness, simultaneously and independently, from a sample. The technique extracts material parameters for both high and low index materials. Second, a fiber-coupled THz system provides an unparalleled opportunity to simulate seismic data collection. We demonstrate the homology between ultra-wideband seismic and THz imaging using multistatic reflective data acquisition. These results broaden the capabilities of THz imaging by borrowing from a mature imaging community. We investigate the resolution limits and show results from both simple and complex layered targets.

Terahertz Time-Domain Spectroscopy (THz-TDS) techniques have been shown over the last decade to be useful in many diverse applications. This thesis describes the implementation of interferometry in imaging with few-cycle terahertz pulses for the purpose of enhancing depth resolution. By configuring terahertz imaging optics in a Michelson interferometric arrangement, a phase shift of approximately Pi radians can be introduced between the interferometer's two arms via the Gouy effect. The resulting destructive interference provides a nearly background-free measurement and a dramatic enhancement in imaging sub-coherence length features. It is possible to image features thinner than 4% of the coherence length of the radiation. This technique could have applications in THz imaging and other THz-TDS systems, as well as in other low-coherence optical tomographic measurements.

We describe observations of the amplitude and phase of an electric field diffusing through a three-dimensional random medium by using Terahertz Time-Domain Spectroscopy. These measurements are spatially resolved with a resolution smaller than the speckle spot size, and temporally resolved with a resolution better than one optical cycle. We propose a new correlation function---time-windowed correlation function---to explore the correlation among the measured waveforms. We have obtained information about individual scattering events experienced by the diffusing fields and how correlation fields evolve temporally. This represents a new method to characterize a multiply scattered wave and opens up new possibilities for imaging in biological media.

The characteristics of a pulsed laser ion source have been studied. A pulse from a heating laser (Nd: YAG laser) desorbs the neutral atoms from the surface of a metal target and then a second pulse or pulses from excimer or dye lasers selectively ionize the desorbed neutrals using the Resonance Ionization Spectroscopy technique. Time-of-flight and electrostatic energy analyzer measurements have been used to study different characteristics of this ion source such as spatial, velocity, and energy distributions. These measurements reveal that although the energy spread of the basic source is relatively high, (FWHM $ sim$ 37 eV) the use of a pulsed acceleration system can reduce this spread by a factor of 5 (to about 7 eV), making the source suitable for collinear laser spectroscopy.

When the iris of a conventional camera is replaced by a mask with multiple apertures, a composite image is formed. Unlike binocular stereopsis, the views from each aperture are superimposed, so that conventional methods in stereo vision do not apply. Still, the local displacement between corresponding points in these views is related to their distance from the camera. This depth cue provides the basis for a new paradigm in passive range sensing--monocular stereopsis. This thesis presents a technique for computing a dense range image from one composite image acquired with a multiple aperture camera. The formation of the composite image is modelled as an echo process, where the depth of a point in the scene is directly related to the spatial delay of its visual echo. Cepstral analysis is the method used to detect this echo. A model of the composite image cepstrum allows measurement of monocular disparity to subpixel precision, as well as an estimate of its associated error distribution. This data, computed over a dense grid, is used to generate a piecewise planar representation of surfaces in the scene, based on a maximum likelihood criterion. Borrowing techniques from visual psychophysics, the spatial resolution of this result is evaluated in terms of an intelligent agent making decisions about its environment. This new range imaging technique is successfully applied to real-world scenes to demonstrate its potential for mobile robot navigation and obstacle avoidance.

This thesis examines the design and characterisation of a compact optical power equalizer module for optical fiber telecommunication applications. The various components of the modules are separately examined in terms of their impact on the overall dimension, performance and loss of the equalizer module. Two components are studied in greater details, namely the micro diffractive optical elements (DOEs), responsible for the multiplexing/demultiplexing of the wavelength channels, and the micromirrors, responsible for creating the beam displacements and thus controlling the attenuation. The DOEs allow for the spatial separation of 64 wavelengths into a two-dimensional array with a compact module. MUMPs surface-micromachined micromirrors have good optical properties: 0.29 dB (93.5%) reflectance, low surface sag and negligible curvature. At operating voltages of less than 15 V, tilts of 16 mrad are obtained, sufficient to provide over 50 dB attenuation for the equalizer. Simulation results show that the diffractive elements have low efficiency due to fabrication limitations. The power loss due to beam clipping is 2.81 dB with overall system efficiency at above 5 dB. Major areas of improvement in the device packaging, assembly and efficiency have been identified for future works.

A dual-wavelength Scanning Near-Field Optical Microscope was developed in order to investigate near-field contrast mechanisms as well as biological samples in air. Using a helium-cadmium laser, light of wavelengths 442 and 325 nanometers is coupled into a single mode optical fiber. The end of the probe is tapered to a sub-wavelength aperture, typically 50 nanometers, and positioned in the near-field of the sample. Light from the aperture is transmitted through the sample and detected in a confocal arrangement by two photomultiplier tubes. The microscope has a lateral topographic resolution of 10 nanometers, a vertical resolution of 0.1 nanometer and an optical resolution of 30 nanometers. Two alternate methods of producing the fiber probes, heating and pulling, or acid etching, are compared and the metal coating layer defining the aperture is discussed. So-called "shear-force" interactions between the tip and sample are used as the feedback mechanism during raster scanning of the sample. An optical and topographic sample standard was developed to calibrate the microscope and extract the ultimate resolution of the instrument. The novel use of two wavelengths enables the authentication of true near-field images, as predicted by various models, as well as the identification of scanning artifacts and the deconvolution of often highly complicated relationships between the topographical and optical images. Most importantly, the use of two wavelengths provides information on the chemical composition of the sample. Areas of a polystyrene film are detected by a significant change in the relative transmission of the two wavelengths with a resolution of 30 nanometers. As a biological application, a preliminary investigation of the composition of Black Spruce wood cell fibers was performed. Comparisons of the two optical channels reveal the expected lignin distributions in the cell wall.