Adaptive Wideband Microwave Signal Processing Using Reconfigurable Integrated Photonics

Abstract

The advancement of wireless communications in domains such as satellites, radars, 5G, and future technologies necessitates the utilization of broadband microwave and millimeter wave signal processing. Unlike their electronic counterparts, photonic systems offer a range of distinct advantages. These include ultra-wide instantaneous bandwidth, agility, ease of dynamic reconfigurability, electromagnetic interference-free operation, reduced footprint, and low losses. In essence, photonic domain processing techniques have been developed to address the limitations and bottlenecks encountered in electronic domain processing.

Integrated microwave photonics (IMWP) emerges as a field that focuses on processing microwave signals in the photonics domain through either monolithic or hybrid integration, allowing for enhanced agility, compactness, and stability. IMWP encompasses a range of functionalities, including filtering, beamforming, instantaneous frequency measurement, generation of arbitrary waveforms, differentiation, time reversal, temporal convolution, and more. By leveraging the advantages of photonics, IMWP offers solutions to overcome challenges and enhance the performance of microwave signal processing, facilitating the requirements of next-generation wireless communication systems.

This thesis focuses on three specific Integrated Microwave Photonics (IMWP) signal processing areas: microwave photonic multi-functional filtering, microwave photonic multi-path self-interference cancellation, and microwave photonic multi-beamforming. The content of the thesis is divided as follows: 20\% dedicated to filtering, 20\% to self-interference cancellation, and 60\% to beamforming.

Microwave Photonic filters (MWPF) play a pivotal role in signal processing by effectively filtering out unwanted portions or noise from input microwave signals. Their versatility extends to various domains, including satellite and wireless communication systems, radar systems, sensors, and radio astronomy. We have developed a novel triple-coupled microring resonators-based multi-functional reconfigurable microwave filter to address the need for advanced filtering capabilities. This innovative filter design offers the unique capability of performing notch, bandpass, and bandstop filtering. The same device can seamlessly switch between different filtering functionalities by dynamically adjusting the coupling coefficients while maintaining high extinction ratios and exceptional out-of-band rejection. The filter's bandwidth can also be tuned flexibly, spanning from hundreds of megahertz to a few gigahertz. Notably, the center frequency is adjustable by precisely controlling the phases of phase shifters strategically placed along the feedback paths of the microring resonators. The practical applications of this reconfigurable filter are extensive. For instance, it can be employed for image rejection in heterodyne receivers or channel selection in channelized receivers. These scenarios demonstrate the filter's ability to enhance receiver performance by effectively mitigating interference and enabling precise frequency selection. Through our novel triple-coupled microring resonators-based multi-functional reconfigurable microwave filter, we have achieved a significant advancement in microwave photonic filtering capabilities, empowering various communication and sensing systems with improved signal quality and enhanced performance.

In-band full duplex (IBFD) wireless communication systems hold tremendous potential for doubling spectral efficiency and data rates in emerging high-capacity networks like 5G and beyond. By enabling simultaneous transmission and reception at the same time and frequency, IBFD technology offers the advantage of enhanced spectral efficiency. However, the presence of in-band self-interference (SI) poses a significant challenge, compromising the signal-to-noise ratio (SNR) unless effectively suppressed by cancellation mechanisms. Thus, the persistent issue of in-band self-interference demands effective mitigation techniques. This thesis focuses on designing and implementing an adaptive multi-path microwave photonic self-interference cancellation architecture specifically engineered for single antenna systems. The proposed canceller strives to achieve substantial cancellation depth across a wide instantaneous bandwidth while ensuring the recovery of the desired signal of interest (SOI) with minimal distortion. It adaptively captures the leaked self-interference signal and multi-path reflections from the surrounding unknown environment, employing adaptive measures to mitigate their impact. This adaptability is achieved through intelligent control and reconfiguration of the multi-tap canceller's phase, time delay, and attenuation. The adaptation process is facilitated by the deployment of the interior-point algorithm, enhancing the cancellation performance.

Multi-beamforming techniques have revolutionized data transmission rates and spectral efficiency by accommodating multiple users and data streams simultaneously. This thesis explores three novel approaches to microwave photonic beamforming. Firstly, "spectral-spatial multi-beamforming" utilizes frequency division multiple access (FDMA) to assign orthogonal frequency channels to individual users. By leveraging multi-wavelength frequency combs, modulated signals are de-multiplexed and directed through programmable high-resolution filters and true-time delay lines for analog beamforming. Our analysis thoroughly evaluated the multi-beamformer's performance, specifically focusing on noise and non-linearity characteristics by assessing the performance metrics such as link gain, noise figure, and spurious free dynamic range.

The second approach, "all-optical NOMA with multi-beamforming," utilizes a single frequency with power domain non-orthogonal multiple access (NOMA) to differentiate users. We observe significant performance improvements by comparing our NOMA system with multi-beamforming to the conventional frequency division multiple access (FDMA) with multi-beamforming. The combination of NOMA and multi-beamforming demonstrates superior performance in terms of spectral efficiency, surpassing the capabilities of FDMA with multi-beamforming.

Lastly, "photonic hybrid beamforming" combines a single frequency with hybrid digital and analog beamforming, reducing hardware complexity while maximizing spectral efficiency. Compared to conventional approaches, our proposed low-complexity partially connected photonic hybrid beamforming technique offers a more efficient and flexible solution for generating multiple analog beams. This leads to improved rate or spectral efficiency, addressing the demand for higher data rates in modern communication systems.