<p>  </p> <p>This thesis introduces a 2.4 GHz hybrid 3-dB coupler with improved out-of-band rejection by incorporating a setup of transverse transmission lines, extensions, and open-circuit stubs at each of the coupler branches. To achieve matching at the design frequency, the <em>ABCD</em> matrix of the above-mentioned setup is calculated and associated to that of a 90° transmission line, leading to exact analytical equations for the impedances and electrical lengths. Simulated coupler design shows a return loss and isolation better than –27 dB as well as an in-band and out-of-band transmission losses better than –4 dB and –25 dB, respectively.</p> <p>The resulting design is applied to a 4×4 Butler matrix, which consists of other components; namely, crossovers, phase shifters and phase matchers. After designing and validating each of these components, they are integrated in the Butler matrix. To validate the radiation (i.e., beamforming) patterns, a microstrip patch antenna array is also designed and verified. The overall network (i.e., Butler matrix, antenna array) is simulated and parameters including phase-differences and 2D and 3D radiation patterns are validated. These electrical parameters are also measured for the fabricated Butler matrix. Based on the 2D and 3D radiation patterns, beamforming is obtained at ±10° and ±30°.</p> <p>Beamforming has applications in many different areas. Some applications like tracking and localization, satellite communication, high power beam-steering, and harmonic radar are also referenced in this thesis.</p>

Recently optical frequency combs and low noise RF tones are drawing increased attention due to applications in spectroscopy, metrology, arbitrary waveform generation, optical signal processing etc. This thesis focuses on the generation of low noise RF tones and stabilized optical frequency combs. The optical frequency combs are generated by a semiconductor based external cavity mode-locked laser with a high finesse intracavity etalon. In order to get the lowest noise and broadest bandwidth from the mode-locked laser, it is critical to know the free spectral range (FSR) of the etalon precisely. First the etalon FSR is measured by using the modified Pound-Drever-Hall (PDH) based method and obtained a resolution of 1 part in 106 , which is 2 order of magnitude better than the standard PDH based method. After optimizing the cavity length, RF driving frequency and PDH cavity locking point, the mode-locked laser had an integrated timing jitter of 3 fs (1 Hz- 100 MHz) which is, to the best of our knowledge, the lowest jitter ever reported from a semiconductor based multigigahertz comb source. The modelocked laser produces ~ 100 comb lines with 10 GHz spacing, a linewidth of ~500 Hz and 75 dB optical signal-to-noise ratio. The same system can also be driven as a regeneratively modelocked laser with greatly improved noise performance. Another way of generating a low noise RF tone is using an opto-electronic oscillator which uses an optical cavity as a high Q element. Due to the harmonic nature of OEOs, a mode selection element is necessary. Standard OEOs use an RF filter having drawbacks such as broad pass band, high loss, and high thermal noise. In our work, a novel optoelectronic scheme which uses an optical filter (Fabry-Perot etalon) as the mode filter instead of an RF filter is demonstrated. This method has the advantage of having ultra-narrow filtering bandwidths ( ~ 10 iv kHz for a 10 GHz FSR and 106 finesse) and an extremely low noise RF signal. Experimental demonstration of the proposed method resulted in a 5-10 dB decrease of the OEO noise compared to the conventional OEO setup. Also, by modifying the etalon-based OEO, and using single side band modulation, an optically tunable optoelectronic oscillator is achieved with 10-20 dB lower noise than dual side band modulation. Noise properties of the OEO as a function of optical frequency detuning is also analyzed theoretically and the results are in agreement with experimental results. The thesis concludes with comments on future work and directions.

A 5.8 GHz RF Power Amplifier (PA) is designed and fabricated in this work, which has very high linearity through a built-in linearizer. The PA is designed, post-layout simulated by Agilent Advanced Design System (ADS) software and fabricated by Win-Semiconductors 0.15µm pHEMT process technology. The post-layout simulation results illustrate the power amplifier can obtained an output power of 23.98 dBm, a power gain of 32.28 dB and a power added efficiency (PAE) of 29% at saturation region, the 3rd intermodulation distortion (IMD3) of -37.7 dBc at 0 dBm input power is attained when operation frequency is 5.8 GHz. We finally obtain that the output power of 17.97 dBm and power gain of 27.97 dB at input power of -10 dBm, PAE of 11.65% at input power of 0 dBm and the IMD3 of -25.66 dBc at -20 dBm input power by measurement, when operation frequency is 5.2 GHz. So the overall RF performance of the PA demonstrates high power, high efficiency and high linearity.

The introduction of Machine Learning (ML) and Deep Learning (DL) techniques into modern radio communications system, a field known as Radio Frequency Machine Learning (RFML), has the potential to provide increased performance and flexibility when compared to traditional signal processing techniques and has broad utility in both the commercial and defense sectors. Existing RFML systems predominately utilize supervised learning solutions in which the training process is performed offline, before deployment, and the learned model remains fixed once deployed. The inflexibility of these systems means that, while they are appropriate for the conditions assumed during offline training, they show limited adaptability to changes in the propagation environment and transmitter/receiver hardware, leading to significant performance degradation. Given the fluidity of modern communication environments, this rigidness has limited the widespread adoption of RFML solutions to date. Transfer Learning (TL) is a means to mitigate such performance degradations by re-using prior knowledge learned from a source domain and task to improve performance on a "similar" target domain and task. However, the benefits of TL have yet to be fully demonstrated and integrated into RFML systems. This dissertation begins by clearly defining the problem space of RF TL through a domain-specific TL taxonomy for RFML that provides common language and terminology with concrete and Radio Frequency (RF)-specific example use- cases. Then, the impacts of the RF domain, characterized by the hardware and channel environment(s), and task, characterized by the application(s) being addressed, on performance are studied, and methods and metrics for predicting and quantifying RF TL performance are examined. In total, this work provides the foundational knowledge to more reliably use TL approaches in RF contexts and opens directions for future work that will improve the robustness and increase the deployability of RFML. / Doctor of Philosophy / The field of Radio Frequency Machine Learning (RFML) introduces Machine Learning (ML) and Deep Learning (DL) techniques into modern radio communications systems, and is expected to be a core component of 6G technologies and beyond. While RFML provides a myriad of benefits over traditional radio communications systems, existing approaches are generally incapable of adapting to changes that will inevitably occur over time, which causes severe performance degradation. Transfer Learning (TL) offers a solution to the inflexibility of current RFML systems, through techniques for re-using and adapting existing models for new, but similar, problems. TL is an approach often used in image and language-based ML/DL systems, but has yet to be commonly used by RFML researchers. This dissertation aims to provide the foundational knowledge necessary to reliably use TL in RFML systems, from the definition and categorization of RF TL techniques to practical guidelines for when to use RF TL in real-world systems. The unique elements of RF TL not present in other modalities are exhaustively studied, and methods and metrics for measuring and predicting RF TL performance are examined.

This thesis work is mainly focused on studying RF performance degradation of a low noise amplifier (LNA) circuit due to hot carrier effect (HCE) in both the weak and strong inversion regions. Since the figures of merit for the RF circuit characterization are gain, noise figure, input, and output matching, the LNA RF performance drift is evaluated in a Cadence SpectreRF simulator subject to these features. This thesis presents hot carrier induced degradation results of an LNA to show that the HCE phenomenon is one of the serious reliability issues in the aggressively scaled RF CMOS design, especially for long-term operation of these devices. The predicted degradation from simulation results can be used design reliable CMOS RF circuits.

No description available.

No description available.

No description available.

No description available.

No description available.