Coherent Multispectral LIDAR in Silicon Photonics using Optical Phased Arrays

点击翻译按钮获取中文摘要

The project will investigate the different aspects of developing a LIDAR sensor on a low-cost, compact silicon chip under the established CMOS fabrication foundry model. The resulting photonic integrated circuit will comprise a large variety of optical components including an on-chip laser, waveguides, metal heaters, amplifiers and photodetectors, but the early focus will be on developing an optical phased array (OPA) to steer and shape the beam in the far-field. In order to have a viable LIDAR sensor with an acceptable system performance using OPAs, certain criteria must be met and satisfied. The potential impact of a fully-integrated LIDAR device on a silicon chip is huge; these devices could replace the bulky LIDAR systems seen on autonomous vehicles and drones up until now at a fraction of the SWaP (Size, weight and power requirements).An optical phased array requires splitting light from a bus waveguide via a multi-mode interference coupler, for example, into an array of waveguides with optical phase shifters attached to each channel to control the phase of each beam separately, and hence steer the beam in free-space in the desired far-field direction. Several issues have to be addressed in order to produce a high quality, narrow divergence output beam via constructive interference of the different beams from each waveguide channel in the array.The first problem is that the phase shifters, operating using the thermo-optic effect, will incur thermal crosstalk issues between adjacent waveguides and disrupt the phases of beams in adjacent channels. This problem can be alleviated by increasing the gap between the waveguides at the location of the shifters. However, this leads to a second problem in that the waveguides will be spaced too far apart at the output of the array, leading to a large beam width in the far-field. At small spacings however the modes between adjacent waveguides will interact and transfer power as in directional waveguide couplers. The waveguides will also have different optical path lengths due to the divergent and convergent path sections. Solutions to all of these problems in order to realise a large optical phased array will be investigated, simulated and then fabricated. The factors will be simulated in Lumerical MODE solutions. There will also be phase nonuniformities induced due to fabrication errors in addition to various other factors. These factors will be monitored and tested post-fabrication to find optimum voltage values for the phase shifters to counter phase nonuniformities. The later parts of the project could involve looking at different wavelengths of lasers for a multispectral analysis of a scene, which would deliver additional target information. Frequency chirping of a continuous wave source will also be studied and its feasibility of on-chip operation compared to using a pulsed source considering factors such as eye-safe operation and efficiency of target detection.