| 摘要: | 相較於脈衝TOF雷達，頻率調變連續波雷達提供獨特的優勢，例如(多個)物體的瞬時速度資訊、消除操作期間的盲點時間以及在工業上有許多的應用，例如汽車或軍事傳感的速度測量、手勢識別 (HGR) 和非接觸式生命體徵監測 (VSM)。特別是VSM的應用，醫療專業人員和醫療保健提供者定期監控生命體徵，具有極高速度靈敏度的FMCW雷達偵測系統非常需要對接收射頻(RF)訊號的小都普勒頻移有最小的相位雜訊影響。根據FMCW雷達計畫中微小的天線尺寸獲得實時4D圖像依然是個挑戰。達成目標最有效率的方法之一是FMCW LiDAR，他將FMCW雷達結構和光電(EO)和電光(OE)轉換模組結合。透過使用內部具有微小光學元件的微型FMCW LiDAR模組可以獲得光學波段1.55um的4-D圖像。此外這些LiDAR圖像通常在方位角和仰角展現出比FMCW雷達有更好的角解析度。此外，當我們的中心頻率從RF提升到光波，還可以期待高速度的靈敏度。商用雷射都普勒測儀器(LDV)已經證明擁有極高速度的靈敏度(~ nm/sec)。儘管如此，相較於由波長掃描雷射驅動的4-D FMCW LiDAR相比，LDV中靜態雷射的干涉訊號無法獲得具有絕對距離資訊的3-D輪廓。使用瞬時線寬與高性能靜止雷射依樣窄的波長掃描雷射，在4-D FMCW LiDAR 和 LDV中達到相對應的速度靈敏度是主要的挑戰。這是因為比較大的線寬會造成測距範圍縮小和對小都普勒頻率偏移的靈敏度降低。為了應對這個挑戰，SILO(自我注入鎖定振盪器)的概念已被改編並且運用在半導體雷射。SILO技術使用低頻的電訊號來穩定雷射的光頻率進而顯著地減少線寬。然而，即使使用SILO，相對於那些FMCW雷達傳輸端的掃瞄RF源，波長掃瞄雷射仍然會遇到更大的相位雜訊和非線性。對於高速度靈敏度的表現，這可能會限制FMCW LiDAR解決小都普勒頻率偏移的能力。為了同時降低 FMCW LiDAR 接收器端的相位和幅度雜訊，我們說明了該 FMCW LiDAR 的接收端如何以極高的速度靈敏度提供高解析度(3-D) + 瞬時速度（4-D）圖像，通過結合自我注入鎖定振盪器 (SILO) 和同時具有高響應度和高飽和電流的高性能雪崩光電二極管 (APD)，分別最小化相位和幅度雜訊。與使用 p-i-n PD 接收器相比，具有 APD 的 LiDAR 系統可以提供更好的 4-D 圖像品質和更高的速度靈敏度，這是由於每個體素對比度的增強。此外，為了進一步提高我們 4-D 圖像中感測速度的靈敏度，我們的掃頻雷射由預編程波形驅動，使光學chirp波形線性化，並在不同時隙（混合波形）分別測量物體的深度和速度。因此，SILO系統產生了更高的速度靈敏度，這導致對於慢速移動（約5µm / sec）物體的4D影像品質比傳統RF接收器參考物體的品質顯著更高。
據我們所知，這裡實現的速度靈敏度（4-D 圖像約為 5 µm/sec）是迄今為止報導的所有 FMCW LiDAR 中最高的，例如在矽光子平台上的慢光光柵結構（75 mm/ 秒，400 毫米/秒），或光子晶體（19 毫米/秒）光束掃描儀和相位多樣性相干檢測（1500 毫米/秒）。由於基於 SILO 的接收端能夠降低在下變頻基帶信號中的相位雜訊，因此微小的都普勒頻移可以非常靈敏地檢測到。我們 FMCW LiDAR 系統中相位和強度雜訊的降低為下一代 4-D LiDAR 開闢了新的可能性
;Compared to pulsed Time-Of-Flight (TOF) radar, Frequency Modulated Continuous Wave (FMCW) radar offers unique advantages, such as instantaneous velocity information of (multiple-) objects, elimination of blind time during operation and has a numerous application in the industry, such as velocity measurement for automotive or military sensing, Hand Gesture Recognition (HGR) and non-contact Vital-Sign Monitoring (VSM). In particular for the VSM application where the vital signs are routinely monitored by medical professionals and health care providers, extremely high velocity sensitivity in FMCW radar detection system is highly desirable with minimum influence of phase noise on the small Doppler frequency shift (< 1 Hz) in the received Radio-Frequency (RF) signals. However, obtaining a real-time 4D (3D + velocity) image based on the FMCW radar scheme with a compact antenna size remains a challenge. One of the effective solutions to achieve this goal is FMCW LiDAR, which combines the FMCW radar architecture with additional Electrical-to-Optical (EO) and Optical-to-Electrical (OE) conversion modules. 4-D images can be obtained at the optical wavelength of 1.55 µm by using a miniaturized FMCW LiDAR module with compact internal optics inside. Moreover, these LiDAR images usually exhibit much better angular resolution in both azimuth and elevation than those of FMCW radar. In addition, when we boost the central frequency from RF to optical wave, high velocity sensitivity can also be expected. Ultra-high velocity sensitivity (~ nm/sec) has been demonstrated in the commercial available laser doppler vibrometers [LDV]. Nevertheless, in contrast to 4-D FMCW LiDAR, which is driven by a wavelength sweeping laser, the 3-D profile with absolute distance information can’t be obtained by the interference signal from a static laser in LDV. The use of a wavelength sweeping laser with an instantaneous linewidth as narrow as that of a high-performance static laser is a major challenge in achieving comparable velocity sensitivity between 4-D FMCW LiDAR and LDV. This is because a larger linewidth can result in a reduced measurement range and lower sensitivity to small Doppler frequency shifts. To address this challenge, the concept of the SILO (Self Injection-Locked Oscillator) has been adapted and applied to semiconductor lasers. The SILO technique uses a low-frequency electrical signal to stabilize the optical frequency of the laser, which can result in a significant reduction in linewidth. However, even with the use of SILO, wavelength sweeping lasers can still suffer from larger phase noise and nonlinearity than those of a swept RF source on the transmitter side of the FMCW radar. This can limit the ability of the FMCW LiDAR to resolve small Doppler frequency shifts for high velocity sensitivity performance. In order to simultaneously reduce the phase and amplitude noises in the FMCW LiDAR receiver end, we illustrate how the receiving end of this FMCW LiDAR can provide high-resolution 3-D + instantaneous velocity (4-D) images with an extremely high velocity sensitivity by combining a self-injection-locked oscillator (SILO) and high-performance avalanche photodiode (APD), which simultaneously has high-responsivity and high saturation current, to minimize the phase and amplitude noise, respectively. Compared with using the p-i-n PD receiver, the LiDAR system with APD one can provide a much better quality of 4-D images with a higher velocity sensitivity due to the enhancements in contrast ratio of each voxel. Besides, to further improve the sensitivity for sensing velocity in our 4-D image, our sweeping laser is driven by a pre-programmed waveform to linearize the optical chirp waveform and separately measure the depth and velocity of object at different time slots (hybrid waveform). Consequently, the SILO system produces a higher velocity sensitivity, resulting in a considerably superior quality of 4-D images of a slow-moving (~ 5µm/sec) object to those of reference one with conventional RF receiver. To the best of our knowledge, the velocity sensitivity achieved here (~ 5 µm/sec with 4-D image) is the highest among all the FMCW LiDARs that have been reported so far such as with slow-light grating structure (75 mm/sec, 400 mm/sec) on silicon photonic platform, or photonic crystal (19 mm/sec) beam scanners and phase-diversity coherent detection (1500 mm/sec).Due to ability of the SILO-based receiving end to reduce phase noise in the down-converted baseband signals, a minor Doppler frequency shift can thus be detected with exceptional sensitivity. The reduction of phase and intensity noise in our FMCW LiDAR system opens up new possibilities for the next generation of 4-D LiDARs. |
