Development, calibration and characterization of silicon photonics based optical phased arrays

Sylvain Guerber,Daivid Fowler,Ismael Charlet,Philippe Grosse,Kim Abdoul-Carime,Jonathan Faugier-Tovar,Bertrand Szelag
DOI: https://doi.org/10.1117/12.2582679
2023-05-05
Abstract:Over the last decade, Optical Phased Arrays (OPA) have been extensively studied, targeting applications such as Light Detection And Ranging (LiDAR) systems, holographic displays, atmospheric monitoring and free space communications. Leveraging the maturity of the silicon photonics platform, the usual mechanical based beam steering system could be replaced by an integrated OPA, significantly reducing the cost and size of the LiDAR while improving its performance (scanning speed, power efficiency, resolution) thanks to solid state beam steering. However, the realization of an OPA that meets the specifications of a LiDAR system (low divergence and single output beam) is not trivial. Targeting the realization of a complete LiDAR system, the technical challenges inherent to the development of high performance OPAs have been studied at CEA LETI. In particular, efficient genetic algorithms have been developed for the calibration of high channel count OPAs as well as an advanced measurement setup compatible with wafer-scale OPA characterization.
Optics,Applied Physics
What problem does this paper attempt to address?
The main objective of this paper is to address the application of optical phased arrays (OPA) in light detection and ranging (LiDAR) systems. Specifically, the paper aims to achieve the following points: 1. **Replace Mechanical Scanning Systems**: Utilize the mature technology of the silicon photonics platform to develop integrated OPAs to replace traditional mechanical scanning systems. This can significantly reduce the size and cost of LiDAR systems while improving their performance (such as scanning speed, power efficiency, and resolution). 2. **High Resolution and Low Divergence**: Achieve a high-performance OPA that meets the requirements of LiDAR systems (low divergence and single output beam). This requires addressing technical challenges, such as how to increase the number and spacing of antennas while ensuring beam quality. 3. **Efficient Calibration Algorithm**: Develop an efficient genetic algorithm to calibrate high-channel-count OPAs. Traditional methods (such as the hill-climbing method) become inefficient as the number of channels increases, whereas genetic algorithms can find the optimal solution faster in large-scale OPAs. 4. **Wafer-Level Characterization System**: Establish a system capable of characterizing OPAs at the wafer level to accelerate the mass production of OPAs. This system includes an automated test platform based on a probe station and achieves large field-of-view imaging through a linear actuator. In summary, the paper aims to promote the practical application of integrated OPAs in LiDAR systems by developing efficient calibration algorithms and advanced characterization systems, thereby advancing technology in fields such as autonomous vehicles.