Integrated lithium niobate microwave photonic processing engine (2306.14415v2)

Abstract: Integrated microwave photonics is an intriguing field that leverages integrated photonic technologies for the generation, transmission, and manipulation of microwave signals in chip-scale optical systems. In particular, ultrafast processing and computation of analog electronic signals in the optical domain with high fidelity and low latency could enable a variety of applications such as MWP filters, microwave signal processing, and image recognition. An ideal photonic platform for achieving these integrated MWP processing tasks shall simultaneously offer an efficient, linear and high-speed electro-optic modulation block to faithfully perform microwave-optic conversion at low power, and a low-loss functional photonic network that can be configured for a variety of signal processing tasks, as well as large-scale, low-cost manufacturability to monolithically integrate the two building blocks on the same chip. In this work, we demonstrate such an integrated MWP processing engine based on a thin-film lithium niobate platform capable of performing multi-purpose processing and computation tasks of analog signals up to 92 giga samples per second at CMOS-compatible voltages. We demonstrate high-speed analog computation, i.e., first- and second-order temporal integration and differentiation with computing accuracies up to 98.1 %, and deploy these functions to showcase three proof-of-concept applications, namely, ordinary differential equation solving, ultra-wideband signal generation and high-speed edge detection of images. We further leverage the image edge detector to enable a photonic-assisted image segmentation model that could effectively outline the boundaries of melanoma lesion in medical diagnostic images, achieving orders of magnitude faster processing speed and lower power consumption than conventional electronic processors.

• The paper introduces a monolithic lithium niobate platform achieving 256 GSa/s processing speed with 98% accuracy in EO modulation.
• It integrates high-speed EO modulation and low-loss photonic structures on a 4-inch wafer, enabling analog computations at up to 67 GHz bandwidth.
• It demonstrates practical applications such as solving ODEs, UWB signal generation, and photonic-assisted medical image edge detection.

Insights into Integrated Lithium Niobate Microwave Photonic Processing

The paper presents a significant advancement in the field of integrated microwave photonics (MWP), leveraging a thin-film lithium niobate (LN) platform to fabricate a high-fidelity, broadband, and low-power microwave signal processing system capable of achieving ultrahigh-speed analog computations. The paper aims to address challenges in ultrahigh-speed signal processing by providing a platform that outperforms traditional MWP systems, especially concerning electro-optic (EO) modulation fidelity, processing speed, and power efficiency.

Core Contributions

The authors detail the development of a monolithically integrated MWP processing engine designed on a 4-inch LN platform. This engine can perform multi-purpose processing and computation tasks for analog microwave signals with processing speeds reaching up to 256 GSa/s at CMOS-compatible voltages. Key contributions include:

• Integration of a high-speed EO modulation block with a low-loss signal processing section on the same chip, fabricated using scalable wafer-scale processes.
• Demonstration of first- and second-order temporal integration and differentiation with bandwidths up to 67 GHz and efficiencies reaching 98.0%.
• Implementation of various signal processing applications, such as ordinary differential equation (ODE) solving, ultra-wideband (UWB) signal generation, and high-speed edge detection for image segmentation.
• Deployment of a photonic-assisted image edge detection model for medical diagnostics, achieving significantly faster processing speeds and lower power consumption than conventional electronic methods.

Technical Significance

The paper underscores the capabilities of the LN platform, particularly highlighting its linear and instantaneous EO response enabled by the Pockels effect, which provides low-loss, broad bandwidths essential for high-fidelity microwave-optic conversion. The integration into a single platform allows the simultaneous use of advanced modulation techniques and low-loss photonic structures, such as ultrahigh-Q microresonators, enhancing the performance of MWP systems.

A notable technical achievement is the platform's ability to process signals with substantial bandwidths that are traditionally challenging for electronic systems. This scalability and high performance make the LN platform an attractive candidate for next-generation communication systems, high-resolution radar, and the Internet of Things, where low-latency and efficient signal processing are paramount.

Quantitative and Experimental Achievements

The platform's capabilities are demonstrated through experiments with impressive results:

• The photonic integration achieved computational accuracies up to 98.0% for fundamental signal processing functions.
• For image processing applications, particularly in medical diagnostics, the system exhibited edge detection accuracies of 96.6% and segmentation accuracies up to 97.3% when tested on medical images.
• The efficiency was further exemplified by the photonic-assisted image segmentation model, which outperformed traditional electronic models in terms of computational speed and energy consumption.

Implications and Future Directions

This work highlights a significant step forward in the integration of LN technologies into MWP systems, potentially shifting the paradigm in how high-speed, low-power signal processing tasks are approached. The demonstration of real-world applications in solving differential equations, generating UWB signals, and medical image processing paves the way for these systems to be part of more extensive telecommunication and data processing infrastructures.

Furthermore, the paper suggests scalable and potentially cost-effective manufacturing methods, an essential consideration for widespread commercial application. Future work could focus on further enhancing the reconfigurability and programmability of these systems, expanding the functional repertoire of LN-based devices, and integrating additional components for even greater MWP functionality. With the rapid technological advancements in AI and machine learning, such photonic systems could significantly contribute to developing efficient, high-speed algorithms tailored for complex computational tasks in a broad context of industrial applications.