利用基因演算法優化矽基側向型接面馬赫詹德行波電光調制器之頻寬

Optimization of bandwidth of Si-Based Lateral Junction Traveling Wave Mach-Zehnder Modulator by using Genetic Algorithm

鍾國方(Kuo-Fang Chung)

指導教授 ： 黃定洧

國立臺灣大學/電機資訊學院/光電工程學研究所/碩士(2019年)

https://doi.org/10.6342/NTU201900765

摘要

隨著近幾年來物聯網、大數據、及雲端計算的蓬勃發展，人們對數據傳輸的速度需求隨之驟增，數據中心 (Datacenter) 及其相關的傳輸能力 (Transmission Capacity) 顯得格外重要。幸運的是，擁有較大傳輸容量及較長傳輸距離的矽光子光互聯非常適合被使用在這些數據中心內與數據中心間的連結上，其利用相容於互補式金屬氧化物半導體 (Complementary Metal Oxide Semiconductor, CMOS) 之成熟的製程技術、低成本、及大量生產的優勢成為數據中心在中短程傳輸上的佼佼者。本篇論文旨在以相容於CMOS之製程技術為基礎、矽光子為背景領域，分析並最佳化光互聯數據發送端 (Transmitter) 中的矽基行波電光調制器 (Si-Based Traveling Wave Optical Modulator)。
本篇論文以IMEC的CMOS製程技術及規範為背景，在眾多不可變動參數的前題下分析多種可影響矽基側向型接面行波馬赫詹德電光調制器 (Si-Based Lateral Junction Traveling Wave Mach-Zehnder Modulator, Si-Based LATWMZM) 之參數、探討其對元件表現所帶來的對立影響、並提出考量射頻損耗之調制效率、插入損耗、及消光比與Lumerical MODE Solution, FDE Solver解出之模態電場數值解及P-N接面於逆偏情形下之一維近似閉合解進行重疊積分，搭配自定義之基因演算法優化當元件操作在光波段1310 nm之調制及射頻表現。在此以 的消光比為優化前題，探討並優化兩種不同射頻信號 命題下的3-dB調制頻寬。優化後之調制頻寬分別為56 GHz與67 GHz、調制效率分別為1.12 V-cm與1.26 V-cm，並以Lumerical Interconnect驗證得出眼圖之消光比大於9 dB。最後，本篇論文以均勻的製程誤差為假設前題並分析其造成的製程容忍度。在製程容忍度分析中，本篇論文參考IMEC提供的製程誤差 (一個標準差大約為7 nm)，分析結果顯示當製程誤差至多到達20 nm時，會分別造成電光3-dB調制頻寬約10 GHz及消光比最多約0.5 dB的誤差值。

關鍵字

數據中心 ； 光互聯 ； 光互聯模組 ； 發送端 ； 矽光子 ； 調制 ； 調制器 ； 光調制器 ； 矽基光調制器 ； 行波電光調制器 ； 調制頻寬 ； 基因演算法

並列摘要

With the bloom of big data, Internet of things, and cloud computing in recent years, the need for the speed of data transmission is increasing. Thus, datacenters and the related transmission capacity have become important. Fortunately, the silicon photonic optical interconnect featuring large transmission capacity and long transmission distance is very suitable for the inter- and intra- datacenter data transmission. Leveraging the advantages of mature fabrication technology, low cost, and mass production from the Complementary Metal Oxide Semiconductor (CMOS) manufacturing technology, the Silicon Photonics platform has become the leading option for the mid- and short- range transmission for the datacenters. Therefore, this thesis aims at the analysis and optimization of the Si-Based Traveling Wave Optical Modulator used in an optical interconnect transmitter based on the CMOS-compatible fabrication technology and the background realm of Silicon Photonics.
With the considerations of the CMOS fabrication technology and the mask rule from IMEC and with the assumption of multiple unchangeable parameters, the thesis analyzed various parameters affecting Si-Based Lateral Junction Traveling Wave Mach-Zehnder Modulator (Si-Based LATWMZM) and discussed various device performances with the use of the modified attenuation-considered figures of merit including modulation efficiency, insertion loss IL, and extinction ratio ER, as well as the overlap integration between the electric field extracted from FDE solver in Lumerical MODE Solution and the analytical carrier density approximation of the reversed-biased P-N junction. In addition, by using the user-defined genetic algorithm, the performance of the device operating at the optical wavelength 1310 nm was optimized. Also, the discussion and the E-O bandwidth optimization of the two different cases with radio-frequency peak-to-peak voltages 2.5 Vpp and 3.5 Vpp were made under the constraints of ER >= 5.5 dB and SEE11 <= -9 dB. The optimization results show that the E-O bandwidth and the modulation efficiency for 2.5 Vpp and 3.5 Vpp are (56 GHz, 1.12 V-cm) and (67 GHz, 1.26 V-cm) respectively. After the optimization, the validations of eye diagrams both with ER > 9 dB were simulated by Lumerical Interconnect. In the last part, the thesis analyzed the tolerance of the device to the fabrication error. The results showed that the possible reductions of BW and ER are about 10 GHz and 0.5 dB respectively for a uniform fabrication error of 20 nm.

並列關鍵字

datacenter ； optical interconnect ； optical interconnect module ； transmitter ； silicon photonics ； modulation ； modulator ； optical modulator ； Si-based modulator ； traveling wave modulator ； modulation bandwidth ； genetic algorithm

參考文獻

參考文獻

1. Data Age 2025: The Evolution of Data to Life-Critical. Don’t Focus on Big Data; Focus on the Data That’s Big. Retrieved on April 05, 2019, from https://www.import.io/wp-content/uploads/2017/04/Seagate-WP-DataAge2025-March-2017.pdf

2. Cisco Global Cloud Index: Forecast and Methodology, 2016–2021 White Paper. (November 19, 2018). Retrieved on April 07, 2019, from https://www.cisco.com/c/en/us/solutions/collateral/service-provider/global-cloud-index-gci/white-paper-c11-738085.html

3. Soref, R., “The past, present, and future of silicon photonics,” IEEE Journal of Selected Topics in Quantum Electronics. 12(6): pp.1678-1687 (2006).

4. Miller, D.A.B., “Device Requirements for Optical Interconnects to Silicon Chips,” Proceedings of the IEEE. 97(7): pp.1166-1185 (2009).