Open Access
Add to BibSonomy
Abstract
Time of flight light detection and ranging (LiDAR) has been tested and used as a key device for auto-driving of vehicles. Frequency-modulated continuous-wave (FMCW) LiDAR potentially achieves a high sensitivity. In this study, we fabricated and tested two components of FMCW LiDAR based on Si photonics. The ranging action was also experimentally simulated. A Si photonic crystal slow light Mach-Zehnder modulator was driven by linearly frequency-chirped signals to generate quasi-frequency-modulated signal light. Then, the light was inserted into a fiber delay line of 20–320 m. Its output was irradiated to a photonic crystal slow beam steering device that acted as an optical antenna via a free-space transmission. The detected light was mixed with the reference light branched after the modulation in balanced photodiodes. A sufficiently sharp beat spectrum was observed, whose frequency well agreed with that expected for the delay line. The experimental simulation of the FMCW LiDAR, thus, was achieved.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Light detection and ranging (LiDAR) has been studied and developed for half a century and used for weather and atmosphere observations as well as three-dimensional (3D) imaging of objects [1–3]. In last decade, various applications of LiDAR have emerged, such as vehicle sensors for advanced driver assistance system (ADAS), robot and drone sensors, fine topographical measurement of geography and buildings, and so on [4–7]. All conventional LiDARs for these applications employ a time-of-flight (TOF) method that uses mechanical rotations of a sensor or a mechanical mirror to scan a laser beam and measures the distance from the roundtrip delay against the object. Such mechanical scanning, however, has issues of large size, high cost of assembly, and low reliability particularly for vibrations of vehicles, robots and drones [8] (first problem). Besides, a high pulse peak power of 50 W or more is usually required to detect sufficient returned light power in the TOF method. Pulsed laser diodes that emit such a high power are available at wavelengths shorter than 1 μm. Its pulse repetition frequency, however, must be moderately suppressed to satisfy eye safety. This issue limits the distance of objects for ranging and the speed of imaging (second problem).
To solve the first problem, non-mechanical optical beam steering devices based on silicon (Si) photonics that enables advanced photonic integration are actively studied and extensively developed. Optical phased arrays [9–13] are particularly attracting attentions as non-mechanical optical beam steering devices, which form an optical beam in an arbitrary direction by controlling the phase of many optical antennas. However, creation of a sharp beam is not simple because of the extremely complicated phase control. A classical approach is to use a waveguide grating that easily forms a sharp beam via diffraction. This approach, however, requires a wide wavelength sweep and/or a large refractive index change of the waveguide to obtain a wide steering angle range [14]. To overcome these constraints, we proposed and demonstrated non-mechanical beam steering devices [15–17] that apply a Si photonic crystal waveguide (PCW) as a diffraction grating and enhance the steering range due to the slow light effect of the PCW.
The frequency-modulated continuous-wave (FMCW) method is a candidate solution for the second problem [18, 19]. FMCW light is radiated toward the objects, and the received reflection light is coherently detected by mixing with the reference light. The distance is measured from a beat frequency between them. Therefore, it potentially achieves higher sensitivity than the TOF method that employs the simple intensity detection of optical pulses. Although the system configuration of the FMCW is much more complicated than TOF, the construction of FMCW can be simplified by exploiting the photonic integration platform of Si photonics. The FMCW does not need a high intensity pulsed laser source. If a telecom wavelength at around 1.55 μm is employed, which is suitable for Si photonics and also beneficial as an eye-safe wavelength, a swept laser source can be used based on distributed feedback laser diodes and sufficient wavelength sweep range and coherency can be achieved [20].
In this study, we experimentally simulated the FMCW ranging action using a Si PCW modulator as well as a PCW optical beam steering device as an optical antenna for light reception. Although the PCW modulator is a compact Mach-Zehnder intensity modulator [21], the quasi-frequency modulation can be obtained by driving it with linearly frequency-chirped electrical signal. In this case, harmonic components and excess beat frequencies occur, but the target beat frequency would be extracted by thresholding and filtering. In this paper, we first describe the reception characteristics of the optical antenna in Section 2. The quasi-frequency modulation is described in Section 3. Finally, in Section 4, we present the ranging action using these two components and a fiber delay line.
2. PCW optical antenna
The PCW optical antenna is schematically shown in Fig. 1. The photonic crystal slab consists of circular holes that are arranged in the Si layer of silicon-on-insulator (SOI) in a triangular lattice. The PCW consists of a line defect in the photonic crystal pattern. All the structure is cladded by SiO2. The guided mode in this waveguide is radiated into free space to form a light beam when some double period is added along the PCW. A large steering angle can be obtained for a slight change in wavelength and/or refractive index of waveguide materials due to the slow light effect in the PCW. As the reverse process, PCW also functions as an optical antenna that receives the light beam arriving from free space at the same angle. As a doubly periodic structure, we have reported a diffraction grating formed on the surface of the SiO2 cladding [15] and a modulation of the hole diameter in the photonic crystal slab [16]. The latter was employed in this study because of its simple fabrication.
The fabrication method is the same as those reported previously [16, 17]. We set the lattice constant of the photonic crystal to be a = 400 nm. The diameter of holes was 2r = 2r0 ± Δr, where 2r0 = 220 nm and Δr = 0, 5, 10, 15 nm. The lattice shift of the third rows of holes for flattening the dispersion characteristics [22] was s3 = 95 nm. The total length of the PCW was L = 1 mm. At the center wavelength of the transmission band, the propagation loss of the PCW without the double periodicity, i.e., Δr = 0 nm, was evaluated to be 30 dB/cm from the output intensities from PCWs of different lengths. The radiation coefficient due to the double periodicity, αrad, was roughly evaluated to be 30, 120, and 220 dB/cm for Δr = 5, 10 and 15 nm, respectively, from the decay of light radiation at near field, which almost agreed to the calculated values [16]. As shown in Fig. 2, the beam angle versus wavelength characteristics, corresponding to the optical antenna gain, were observed for Δr = 5 nm. For the radiation characteristics as shown in Fig. 2(a), wavelength-tunable laser light was coupled from a fiber-lens module to the Si wire waveguide via the spot size converter (SSC) and radiated from the PCW. A far-field pattern (FFP) was received by a far-field microscope and an InGaAs camera (Raptor, OWL1280 BIS-SWR, SXGA, pixel pitch 10 μm) with a high angular resolution of 0.0288°. The angle gradually decreased by increasing the wavelength. Because the beam angle varies sensitively to the wavelength due to the slow light effect in the PCW, the wavelength sensitivity in this experiment was 0.95°/nm, which is approximately 7 times larger than that of simple diffraction gratings. For the reception characteristics as shown in Fig. 2(b), the free-space light beam was obtained from the laser source and a fiber collimator. The light was irradiated to the optical antenna with a tilt angle θ taken from the direction normal to the device surface. Coupled light was extracted to the fiber and evaluated by an optical power meter. The appearance of light irradiation is also shown in Fig. 3(a). The reception wavelength was shifted by changing the beam angle. The wavelength sensitivity was 0.95°/nm, which agrees precisely with the radiation one.
Fig. 3 Reception of free-space beam irradiated to the PCW antenna at θ = 24°. (a) Appearance of light reception. The right half shows a PCW array, while the left, the SSCs at the chip facet. (b) Relative intensity spectrum of received light where light output from the laser source is used as a reference.
Figure 3(b) shows the detail of the received power spectrum at θ = 24°. A sharp peak appeared at 1533 nm, indicating that the used PCW optical antenna possessed large angular dispersion. The full-width at half-maximum of the peak, δλ, was 0.15 nm. Because the wavelength sensitivity of the beam angle was Δθ/Δλ = 0.95°/nm, the above δλ corresponds to δθ = 0.14°.
As shown in Fig. 3(a), the light spot irradiated the PCW with a 600 μm effective length (right side of this figure). The effective modal width of the PCW was calculated to be ~1 μm. Then, the couplable area of one PCW was 600 μm2. Because the area of the irradiation spot was π(500 × 483) μm2 = 7.6 × 105 μm2, the coupling loss that was simply estimated from the area overlap was 31 dB. More precisely, we have to count the modal mismatch between the collimated beam and PCW. If we assume that the calculated modal profile in the lateral direction and the exponential decay in the longitudinal direction, as well as that due to the upward and downward symmetry produced a 13 dB loss, the received intensity in Fig. 3(b) can be explained by further considering the transmission loss of 3 dB from the PCW to the fiber-lens module via the SSC and 3 dB loss in fiber-related components. If a cylindrical lens is placed above the PCW [18] and its focusing is ideal, the area overlap loss of 31 dB and a lateral modal mismatch loss of roughly 4 dB would be eliminated. Furthermore, a vertical asymmetric structure such as that reported in [23] would reduce the loss by ~4 dB. In addition, a modulation of the radiation coefficient of the PCW flattens the amount of radiation and reduces the longitudinal mismatch loss by ~4 dB.
3. Quasi-FM modulation
In the FMCW system, the frequency of light is chirped temporally by a sawtooth function as an example. Then, a frequency difference occurs between reference light and delayed signal light that is reflected back from an object. A beat frequency corresponding to the delay is generated by mixing these two lights. The delay and the corresponding free-space distance to the object are measured from this frequency. The simplest frequency chirp can be achieved by the wavelength shift of the laser source, for example, by the carrier plasma effect. In this case, however, some nonlinear chirp cannot be avoided, resulting in the broadening of the beat spectrum and degradation of the range resolution. Some interferometric methods such as K clock correction are required to compensate for this, which make the system complicated [19]. In this study, the frequency chirp was realized not by such compensation but by the simple modulation using highly linear electrical drive signals.
The used PCW modulator (MOD) is shown in Fig. 4. This device was fabricated by the same process as that for the optical antenna having the same structural parameters except for the phase shifter length of L = 300 μm. A wavy p-n junction having a 600 nm width and a 600 nm period was formed in the PCW [21]. An arbitrary waveform generator (AWG, Keysight M8195A, sampling rate 65 GS/s) was used for generating the drive signal. Sinusoidal signal at a frequency f = 500 MHz and the corresponding spectrum exhibiting two sidebands are shown in Fig. 5(a). Because the push-pull drive was employed, the carrier frequency in-between the sidebands was well suppressed. Figure 5(c) shows the sawtooth frequency chirping having a period of T = 10 μs and a range of f = 4–5 GHz (modulation bandwidth B = 1 GHz), appearing the almost linear frequency shift in the far view. The time-averaged spectrum of one sideband is shown in Fig. 5(d). The one including harmonic components produced by the nonlinear response of the Mach-Zehnder modulator is shown in Fig. 5(e). These harmonics may become noise in the FMCW. Generally, however, the beat signal arising from the harmonics is much weaker than that from the sidebands, and can be eliminated by thresholding. Furthermore, if the lowest frequency of the drive signal is set higher than B, all frequency components of modulated light and unwanted beat frequencies are separated and eliminated by RF filtering.
Fig. 5 Example of modulated signal. (a) Sinusoidal wave from AWG. (b) Spectrum of modulated light. (c) Temporal frequency of chirped signal. (d) Spectrum of frequency-chirped sideband. (e) Wide-range spectrum of frequency-chirped signal.
4. Ranging action
The ranging using the FMCW was experimentally simulated using the measurement setup as shown in Fig. 6, where the roundtrip light between objects was replaced with light guided in a fiber delay line of various lengths. Light from a wavelength-tunable laser source (Santec TSL-550) was modulated by the PCW modulator having the sawtooth chirp signal of f = 0.5–1.0 GHz (B = 500 MHz), T = 10 μs, Vpp = 5 V, and the bias voltage VDC = 3 V. If the initial phase difference between two arms of the modulator, ϕ, is set at π/2 and Vpp is sufficiently small, the harmonic components are suppressed. In this experiment, however, the harmonic components remained due to the relatively large Vpp. In addition, the final signal takes a minimum value at this ϕ when the quasi-FM modulation with double sidebands is used [24]. Consequently, we investigated an optimum ϕ that balanced a large sidelobe intensity and small carrier and harmonic components. The output light was amplified by an erbium-doped fiber amplifier (EDFA). The amplified spontaneous emission was filtered by a tunable bandpass filter (BPF). Thereafter, the light was branched. The one was guided through the fiber delay line of the length Ldelay and irradiated to the PCW optical antenna after the free-space propagation from the fiber collimator. The received light was finally coupled into the fiber-lens module. Then, it was mixed with the reference one after the polarization control and detected by balanced photodiodes (BPDs) having a 10 GHz bandwidth. The beat frequency was observed by a spectrum analyzer. The linewidth of used laser source was Δν = 70 kHz and the corresponding coherence length was Lc = c/nΔν ≈3 km for a fiber index n = 1.45. The coherent detection is possible when Ldelay < Lc. The beat frequency is given by
where Δf is the frequency shift due to delays other than that in the fiber delay line. The measured beat spectra are shown in Fig. 7(a). The fbeat shifted linearly with Ldelay. The theoretical values almost fit to experimental ones when assuming Δf = 1 MHz. Some spectra arising from harmonic components were also observed. The thresholding, however, was easy, because all of them were lower than the main peak by more than 30 dB.Fig. 7 Demonstration of ranging action. (a) Beat spectra for various fiber lengths. (b) Magnified view of beat spectrum.
Figure 7(b) shows the magnified view of the main peak. Comb-like spectra whose spacing T−1 = 100 kHz were observed. Because one of these spectra was converted to a distance, the range resolution was given by the well-known relation
For B = 500 MHz and n = 1.45 in this experiment, ΔLdelay equaled to 41 cm (In actual LiDARs, ΔLdelay would be 30 cm for n ≈1 in air). The peak intensity was 60 dB higher than the background at Ldelay = 20 m, but it decreased to 30 dB at Ldelay = 320 m with expanded peak width. Because the loss in the fiber was negligible, they were likely to be caused by the decrease in coherence, although Ldelay < Lc was satisfied.5. Conclusions
We simulated experimentally the ranging action of the FMCW LiDAR based on Si photonics and photonic crystal slow light waveguides. The doubly-periodic PCW beam steering device, which achieved high wavelength sensitivity due to the slow light effect, was used as an optical antenna for receiving free-space light beam that satisfied the resonance condition. If it was used for LiDARs, this condition worked effectively for eliminating unwanted background light input. We used a compact PCW modulator driven by the frequency-chirped signal for the quasi-FM modulation of light. A high linearity of the chirping was confirmed from a sharp spectrum of the beat signal without compensation. The ranging was performed for a distance up to 320 m. Although it used the condition that the loss was almost neglected by using a fiber delay line, it is a promising result that suggests the practical application of this device to LiDAR if the light reception efficiency is improved by adding a collimation lens, introducing a unidirectional PCW antenna structure and modulated radiation coefficient, and reducing parasitic losses.
Funding
Accelerated Innovation Research Initiative Turning Top Science and Ideas into High-Impact Values (ACCEL); Japan Science and Technology Agency (JST) (JPMJAC1603).
References and links
1. E. E. Uthe, “Lidar evaluation of smoke and dust clouds,” Appl. Opt. 20(9), 1503–1510 (1981). [CrossRef] [PubMed]
2. H. Okamoto, K. Sato, T. Nishizawa, N. Sugimoto, T. Makino, Y. Jin, A. Shimizu, T. Takano, and M. Fujikawa, “Development of a multiple-field-of-view multiple-scattering polarization lidar: comparison with cloud radar,” Opt. Express 24(26), 30053–30067 (2016). [CrossRef] [PubMed]
3. T. Hakala, J. Suomalainen, S. Kaasalainen, and Y. Chen, “Full waveform hyperspectral LiDAR for terrestrial laser scanning,” Opt. Express 20(7), 7119–7127 (2012). [CrossRef] [PubMed]
4. W. D. Jones, “Keeping cars from crashing,” IEEE Spectr. 38(9), 40–45 (2001). [CrossRef]
5. U. Weiss and P. Biber, “Plant detection and mapping for agricultural robots using a 3D LIDAR sensor,” Robot. Auton. Syst. 59(5), 265–273 (2011). [CrossRef]
6. T. T. Sankey, J. McVay, T. L. Swetnam, M. P. McClaran, P. Heilman, and M. Nichols, “UAV hyperspectral and lidar data and their fusion for arid and semi-arid land vegetation monitoring,” Remote Sensing in Ecology and Conservation 4(1), 20–33 (2018). [CrossRef]
7. G. Shao, G. Shao, J. Gallion, M. R. Saunders, J. R. Frankenberger, and S. Fei, “Improving Lidar-based aboveground biomass estimation of temperate hardwood forest with varying site productivity,” Remote Sens. Environ. 204, 872–882 (2018). [CrossRef]
8. I. Puente, H. Gonzalez-Jorge, J. Martinez-Sanchez, and P. Arias, “Review of mobile mapping and surveying technologies,” Measurement 46(7), 2127–2145 (2013). [CrossRef]
9. K. Van Acoleyen, H. Rogier, and R. Baets, “Two-dimensional optical phased array antenna on silicon-on-insulator,” Opt. Express 18(13), 13655–13660 (2010). [CrossRef] [PubMed]
10. J. K. Doylend, M. J. R. Heck, J. T. Bovington, J. D. Peters, L. A. Coldren, and J. E. Bowers, “Two-dimensional free-space beam steering with an optical phased array on silicon-on-insulator,” Opt. Express 19(22), 21595–21604 (2011). [CrossRef] [PubMed]
11. J. Sun, E. Timurdogan, A. Yaacobi, E. S. Hosseini, and M. R. Watts, “Large-scale nanophotonic phased array,” Nature 493(7431), 195–199 (2013). [CrossRef] [PubMed]
12. H. Abediasl and H. Hashemi, “Monolithic optical phased-array transceiver in a standard SOI CMOS process,” Opt. Express 23(5), 6509–6519 (2015). [CrossRef] [PubMed]
13. D. N. Hutchison, J. Sun, J. K. Doylend, R. Kumar, J. Heck, W. Kim, C. T. Phare, A. Feshali, and H. Rong, “High-resolution aliasing-free optical beam steering,” Optica 3(8), 887 (2016). [CrossRef]
14. H. Hashiguchi, K. Kondo, T. Baba, and H. Arai, “An optical leaky wave antenna by waffled structure,” J. Lightwave Technol. 35(11), 2273–2279 (2017). [CrossRef]
15. K. Kondo, T. Tatebe, S. Hachuda, H. Abe, F. Koyama, and T. Baba, “Fan-beam steering device using a photonic crystal slow-light waveguide with surface diffraction grating,” Opt. Lett. 42(23), 4990–4993 (2017). [CrossRef] [PubMed]
16. H. Abe, M. Takeuchi, G. Takeuchi, H. Ito, T. Yokokawa, K. Kondo, Y. Furukado, and T. Baba, “Two-dimensional beam-steering device using a doubly periodic Si photonic-crystal waveguide,” Opt. Express 26(8), 9389–9397 (2018). [CrossRef] [PubMed]
17. G. Takeuchi, Y. Terada, M. Takeuchi, H. Abe, H. Ito, and T. Baba, “Thermally controlled Si photonic crystal slow light waveguide beam steering device,” Opt. Express 26(9), 11529–11537 (2018). [CrossRef] [PubMed]
18. H. Abe, and T. Baba, “Theoretical performance of FMCE LiDAR using Si photonics and photonic crystal waveguides,” Cong. Int. Commission for Opt., Tu1A–04 (2017).
19. C. V. Poulton, A. Yaacobi, D. B. Cole, M. J. Byrd, M. Raval, D. Vermeulen, and M. R. Watts, “Coherent solid-state LIDAR with silicon photonic optical phased arrays,” Opt. Lett. 42(20), 4091–4094 (2017). [CrossRef] [PubMed]
20. T. DiLazaro and G. Nehmetallah, “Multi-terahertz frequency sweeps for high-resolution, frequency-modulated continuous wave ladar using a distributed feedback laser array,” Opt. Express 25(3), 2327–2340 (2017). [CrossRef] [PubMed]
21. Y. Terada, T. Tatebe, Y. Hinakura, and T. Baba, “Si Photonic Crystal Slow-Light Modulators with periodic p–n junctions,” J. Lightwave Technol. 35(9), 1684–1692 (2017). [CrossRef]
22. T. Tamura, K. Kondo, Y. Terada, Y. Hinakura, N. Ishikura, and T. Baba, “Silica-Clad Silicon Photonic Crystal Waveguides for Wideband Dispersion-Free Slow Light,” J. Lightwave Technol. 33(14), 3034–3040 (2015).
23. G. Roelkens, D. Van Thourhout, and R. Baets, “High efficiency grating coupler between silicon-on-insulator waveguides and perfectly vertical optical fibers,” Opt. Lett. 32(11), 1495–1497 (2007). [CrossRef] [PubMed]
24. F. Aflatouni, B. Abiri, A. Rekhi, and A. Hajimiri, “Nanophotonic coherent imager,” Opt. Express 23(4), 5117–5125 (2015). [CrossRef] [PubMed]
