Fingerprints have been widely used as a practical method of biometrics authentication or identification with a significant level of security. However, several spoofing methods have been used in the last few years to bypass fingerprint scanners, thus compromising data security. The most common attacks occur by the use of fake fingerprint during image capturing. Imposters can build a fake fingerprint from a latent fingerprint left on items such as glasses, doorknobs, glossy paper, etc. Current mobile fingerprint scanning technology is incapable of differentiating real from artificial fingers made from gelatin molds and other materials. In this work, the adequacy of terahertz imaging was studied as an alternative fingerprint scanning technique that will enhance biometrics security by identifying superficial skin traits. Terahertz waves (0.1 – 10 THz) are a non-ionizing radiation with significant penetration depth in several non-metallic materials. Several finger skin features, such as valley depth and sweat ducts, can possibly be imaged by employing the necessary imaging topology. As such, two imaging approaches 1) using quasi-optical components and 2) using near-field probing were investigated. The numerical study is accomplished using a commercial Finite Element Method tool (ANSYS, HFSS) and several laboratory experiments are conducted to evaluate the imaging performance of the topologies. The study has shown that terahertz waves can provide high spatial resolution images of the skin undulations (valleys and ridges) and under certain conditions identify the sweat duct pattern.

Articially engineered two-dimensional materials, which are widely known as

metasurfaces, are employed as ground planes in various antenna applications. Due to

their nature to exhibit desirable electromagnetic behavior, they are also used to design

waveguiding structures, absorbers, frequency selective surfaces, angular-independent

surfaces, etc. Metasurfaces usually consist of electrically small conductive planar

patches arranged in a periodic array on a dielectric covered ground plane. Holographic

Articial Impedance Surfaces (HAISs) are one such metasurfaces that are capable of

forming a pencil beam in a desired direction, when excited with surface waves. HAISs

are inhomogeneous surfaces that are designed by modulating its surface impedance.

This surface impedance modulation creates a periodical discontinuity that enables a

part of the surface waves to leak out into the free space leading to far-eld radia-

tion. The surface impedance modulation is based on the holographic principle. This

dissertation is concentrated on designing HAISs with

Desired polarization for the pencil beam

Enhanced bandwidth

Frequency scanning

Conformity to curved surfaces

HAIS designs considered in this work include both one and two dimensional mod-

ulations. All the designs and analyses are supported by mathematical models and

HFSS simulations.

Kinetic inductance springs from the inertia of charged mobile carriers in alternating electric fields and it is fundamentally different from the magnetic inductance which is only a geometry dependent property. The magnetic inductance is proportional to the volume occupied by the electric and magnetic fields and is often limited by the number of turns of the coil. Kinetic inductance on the other hand is inversely proportional to the density of electrons or holes that exert inertia, the unit mass of the charge carriers and the momentum relaxation time of these charge carriers, all of which can be varied merely by modifying the material properties. Highly sensitive and broadband signal amplifiers often broaden the field of study in astrophysics. Quantum-noise limited travelling wave kinetic inductance parametric amplifiers offer a noise figure of around 0.5 K ± 0.3 K as compared to 20 K in HEMT signal amplifiers and can be designed to operate to cover the entire W-band (75 GHz – 115 GHz).The research cumulating to this thesis involves applying and exploiting kinetic inductance properties in designing a W-band orthogonal mode transducer, quadratic gain phase shifter with a gain of ~49 dB over a meter of microstrip transmission line. The phase shifter will help in measuring the maximum amount of phase shift ∆ϕ_max (I) that can be obtained from half a meter transmission line which helps in predicting the gain of a travelling wave parametric amplifier. In another project, a microstrip to slot line transition is designed and optimized to operate at 150 GHz and 220 GHz frequencies, that is used as a part of horn antenna coupled microwave kinetic inductance detector proposed to operate from 138 GHz to 250 GHz. In the final project, kinetic inductance in a 2D electron gas (2DEG) is explored by design, simulation, fabrication and experimentation. A transmission line model of a 2DEG proposed by Burke (1999), is simulated and verified experimentally by fabricating a capacitvely coupled 2DEG mesa structure. Low temperature experiments were done at 77 K and 10 K with photo-doping the 2DEG. A circuit model of a 2DEG coupled co-planar waveguide model is also proposed and simulated.

This work implements three switched mode power amplifier topologies namely inverse class-D (CMCD), push-pull class-E and inverse push-pull class-E, in a GaN-on-Si process for medium power level (5-10W) femto/pico-cells base-station applications. The presented power amplifiers address practical implementation design constraints and explore the fundamental performance limitations of switched-mode power amplifiers for cellular band. The designs are analyzed and compared with respect to non-idealities like finite on-resistance, finite-Q of inductors, bond-wire effects, input signal duty cycle, and supply and component variations. These architectures are designed for non-constant envelope inputs in the form of digitally modulated signals such as RFPWM, which undergo duty cycle variation. After comparing the three topologies, this work concludes that the inverse push-pull class-E power amplifier shows lower efficiency degradation at reduced duty cycles. For GaN based discrete power amplifiers which have less drain capacitance compared to GaAs or CMOS and where the switch loss is dominated by wire-bonds, an inverse push-pull class-E gives highest output power at highest efficiency. Push-pull class-E can give efficiencies comparable to inverse push-pull class-E in presence of bondwires on tuning the Zero-Voltage Switching (ZVS) network components but at a lower output power. Current-Mode Class-D (CMCD) is affected most by the presence of bondwires and gives least output power and efficiency compared to other two topologies. For systems dominated by drain capacitance loss or which has no bondwires, the CMCD and push-pull class-E gives better output power than inverse push-pull class-E. However, CMCD is more suitable for high breakdown voltage process.