Selected Research Projects
A Machine Learning Approach to Automated Non-Foster Circuit Synthesis (Sponsored by DARPA)
Due to the complexity and multidimensional variables of a non-Foster circuit, it is nearly impossible to design a realistic non-Foster circuit analytically while meeting the requirements of negative impedance, stability, efficiency, and bandwidth, etc. A machine-learning-based approach provides a promising solution for such a problem, according to the universal approximation theorem. This research project builds upon the latest advancements of the two different fields of machine learning and non-Foster circuit design. It has the potential to revitalize the fields of circuit synthesis, metamaterials, and electrically-small antennas.
Intelligent 5G Hardware (Sponsored by UIdaho ORED RISE)
One method to achieve in-band full-duplex operation is to design an antenna that can simultaneously transmit and receive (STAR) signals at the same frequency. Such STAR antennas, along with interference cancellation circuits and signal processing algorithms, allow in-band full-duplex radios and 5G cellular networks to achieve extremely high isolation (e.g., 120 dB). Our research in this area is currently focused on the design of wideband STAR antennas and compact and ultra-low profile STAR antennas that do not require complicated self-cancellation circuits or feed networks.
Electrically-small antennas (ESAs) can be passively matched only over very narrow bandwidths and the resulting antennas have low gains. These are major limiting factors for ESAs used in transmit applications, especially at HF and lower VHF frequency bands. In one of our recent projects, we discussed the challenges of transmit ESA matching circuit design and the design process of a new non-Foster transmit matching architecture for electrically-small monopole antennas that achieves wide bandwidth, high transmission efficiency (transducer power gain), and stability at the same time. The proposed circuit is composed of a current buffer (for high isolation), a transformer (for real-part matching), and a negative impedance converter (for imaginary-part matching). The measured -6 dB (-10 dB) S11 fractional bandwidth of the proposed non-Foster transmitting system is 110% (39%), while the maximum bandwidth that can be achieved is 0.076% (0.047%) when matched with conventional passive matching. The transmission efficiency of the system is improved by as much as 34.4 dB compared to the same antenna without the proposed non-Foster matching circuit, and it retains an enhanced efficiency over the entire frequency band of operation (26 MHz–89 MHz). The system remains stable within this frequency band. The measurement results of the compact and broadband transmitting antenna prototype verify the design concept.
We are further exploring the non-Foster area, with emphasis on high power, stability, wide bandwidth, and high efficiency.
Many high-frequency (HF) antennas have significantly smaller dimensions than the wavelength at which they operate and thus suffer from narrow bandwidths. In many military applications, such HF antennas are mounted on relatively large metallic platforms. In one of our recent projects, we examined how a platform-mounted antenna can be used to excite the characteristic modes (CMs) of the platform itself to increase the overall bandwidth of the system. In this case, the platform will act as the main radiator and the mounted antennas act primarily as the coupling mechanism between the antenna and the external circuit. We use the theory of CMs to identify the appropriate platform modes and determine the practical means of exciting them. This allows for significantly increasing the bandwidth of the antenna system compared to the bandwidth of the antenna system in isolation. This approach is employed to enhance the bandwidth of a horizontally polarized HF loop antenna system by as much as 10 times compared to that of a stand-alone full-loop antenna. Scaled models of the proposed antennas were fabricated and experimentally characterized. Measurement results are in good agreement with the theoretically predicted results and demonstrate the feasibility of using the proposed approach in designing bandwidth-enhanced platform-mounted HF antennas.
We are applying this bandwidth enhancement method to various applications, in combination with several other techniques, to achieve the desired characteristics.
