[Paper Review] Chaotic microcomb inertia-free parallel ranging

Anton Lukashchuk, Johann Riemensberger+3 more|arXiv (Cornell University)|Dec 29, 2022

Advanced Fiber Laser Technologies4 citations

TL;DR

This paper presents a chaotic microcomb-based, inertia-free 2D parallel LiDAR system that enables passive beam steering using a microresonator frequency comb operated in the chaotic modulation instability regime. By combining a VIPA and diffraction grating for 2D spectral dispersion, 40 independent comb lines provide parallel, wavelength-multiplexed ranging with 20 kS/s sampling per channel and 200 fps frame rates, achieving megapixel-level acquisition speeds without mechanical scanning.

ABSTRACT

Ever growing pixel acquisition rates in the fields of augmented reality, autonomous driving and robotics have increased interest in solid state beam scanning without moving parts. Modern photonic integrated laser ranging advances towards passive beam steering solutions. Recently demonstrated imagers based on focal plane arrays, nanophotonic metasurfaces, and optical phased arrays enable unprecedented pixel resolution and measurement speed. However, parallelization of >100 lasers and detectors - successfully implemented in commercial time-of-flight sensors - has not been widely adopted for passive scanning approaches. Here, we show both inertia-free and parallel light detection and ranging (LiDAR) with microresonator frequency combs. We used 40 independent channels of a continuously scanned microresonator frequency comb operated in the chaotic regime in combination with optical dispersive elements to perform random modulation LiDAR with 2D passive beam steering.

Motivation & Objective

• To develop a solid-state, inertia-free LiDAR system capable of high-speed 2D beam scanning without moving parts.
• To overcome the limitation of existing passive beam steering methods by enabling true 2D parallel ranging using a single microcomb source.
• To demonstrate parallel detection of 40 independent LiDAR channels via wavelength-division multiplexing and cross-correlation ranging.
• To achieve high frame rates (up to 200 fps) and megapixel-level sampling using chaotic microcombs and passive optical dispersers.
• To enable scalable, chip-integrated LiDAR for applications in autonomous vehicles, robotics, and augmented reality.

Proposed method

• A continuous-wave pump laser drives a Si3N4 microresonator to generate a chaotic frequency comb via modulation instability, producing 40 independent, broadband, noisy comb lines for random modulation LiDAR.
• The chaotic comb is spectrally dispersed in 2D using a combination of a virtually imaged phased array (VIPA) with 30 GHz free spectral range (FSR) and a diffraction grating, mapping each comb line to a unique spatial angle.
• Simultaneous piezo and heater actuation of the laser and microresonator enables 30 GHz frequency scanning at 100 Hz, maintaining pump-cavity detuning lock during scanning.
• Each comb line is individually accessed via wavelength-division multiplexers and detected in parallel using fiber-coupled photoreceivers with pre-amplification to overcome speckle-induced intensity loss.
• Range is determined via cross-correlation between the reference signal and the backscattered return, with time delay estimation providing distance resolution.
• The system leverages the high finesse (F ≈ 100) and angular dispersion of the VIPA to achieve fine angular resolution and large field of view (1.9° × 7.7°).

Experimental results

Research questions

• RQ1 Can chaotic microcombs enable passive, inertia-free 2D beam steering in parallel LiDAR systems?
• RQ2 How can a single microresonator comb source support parallel, wavelength-multiplexed ranging across multiple independent channels?
• RQ3 What is the maximum achievable frame rate and sampling rate for such a system under practical scanning and detection constraints?
• RQ4 How does the combination of VIPA and diffraction grating enable effective 2D angular mapping without mechanical motion?
• RQ5 What are the key limitations in angular resolution, signal-to-noise ratio, and system scalability due to frequency noise and speckle effects?

Key findings

• The system demonstrated 40 independent, parallel LiDAR channels using a single chaotic microcomb source, each with individual wavelength-multiplexed detection.
• A 30 GHz frequency scan at 100 Hz was achieved via synchronized piezo and heater control, enabling full vertical scanning of the VIPA FSR.
• The sampling rate per comb channel reached 20 kS/s with a 50 µs pixel acquisition time, enabling 200 fps frame rates when using 50 channels.
• The system achieved a total of 50 × 100 = 5,000 pixels per frame, with potential for megapixel-level sampling by increasing the number of comb lines.
• The field of view was measured at 1.9° × 7.7°, with angular dispersion limited by VIPA FSR and finesse (F ≈ 100).
• Despite speckle-induced intensity fluctuations, the use of a pre-amplifier and retroreflector enabled detection above the digitizer noise floor (5 mV).