Vertical-cavity surface-emitting lasers (VCSELs) are the ideal optical sources for data communication and sensing. In data communication, large data rates combined with excellent energy efficiency and temperature stability have been achieved based on advanced device design and modulation formats. VCSELs are also promising sources for photonic integrated circuits due to their small footprint and low power consumption. Also, VCSELs are commonly used for a wide variety of applications in the consumer electronics market. These applications range from laser mice to three-dimensional (3D) sensing and imaging, including various 3D movement detections, such as gesture recognition or face recognition. Novel VCSEL types will include metastructures, exhibiting additional unique properties, of largest importance for next-generation data communication, sensing, and photonic integrated circuits.

https://depositonce.tu-berlin.de/bitstream/11303/13913/1/Liu_etal_2021.pdf

Vertical-cavity surface-emitting lasers for data communication and sensing

GaAs-based 850 nm VCSELs are the standard light source in short-reach high-speed optical interconnects in datacenters and supercomputers. The last few years have seen an impressive increase in VCSEL speed, reaching modulation bandwidths of 28 GHz and single-channel data rates of 57 Gbit/s (non-equalized) and 71 Gbit/s (equalized) under direct NRZ modulation [1-3]. Future optical interconnects will require even faster VCSELs capable of higher-speed transmission at lower energy dissipation. Here we present the results from our new generation of high-speed 850 nm VCSELs, a further development of our previous VCSEL design [1-3].

https://publications.lib.chalmers.se/records/fulltext/231003/local_231003.pdf

High speed 850 nm VCSEL with 30 GHz modulation bandwidth

Client-side optics are facing an ever-increasing upgrading pace, driven by upcoming 5G related services and datacenter applications. The demand for a single lane data rate is soon approaching 200 Gbps. To meet such high-speed requirement, all segments of traditional intensity modulation direct detection (IM/DD) technologies are being challenged. The characteristics of electrical and optoelectronic components and the performance of modulation, coding, and digital signal processing (DSP) techniques are being stretched to their limits. In this context, we witnessed technological breakthroughs in several aspects, including development of broadband devices, novel modulation formats and coding, and high-performance DSP algorithms for the past few years. A great momentum has been accumulated to overcome the aforementioned challenges. In this article, we focus on IM/DD transmissions, and provide an overview of recent research and development efforts on key enabling technologies for 200 Gbps per lane and beyond. Our recent demonstrations of 200 Gbps short-reach transmissions with 4-level pulse amplitude modulation (PAM) and discrete multitone signals are also presented as examples to show the system requirements in terms of device characteristics and DSP performance. Apart from digital coherent technologies and advanced direct detection systems, such as Stokes–vector and Kramers–Kronig schemes, we expect high-speed IM/DD systems will remain advantageous in terms of system cost, power consumption, and footprint for short reach applications in the short- to mid- term perspective.

/pdf/200-gbps-lane-im-dd-technologies-for-short-reach-optical-2nuta7bqsd.pdf

200 Gbps/Lane IM/DD Technologies for Short Reach Optical Interconnects

The estimation of the quality of transmission (QoT) in optical systems with machine learning (ML) has recently been the focus of a large body of research. We discuss the sources of inaccuracy in QoT estimation in general; we propose a taxonomy for ML-aided QoT estimation; we briefly review ML-aided optical performance monitoring, a tightly related topic; and we review and compare all recently published ML-aided QoT articles.

Machine learning techniques for quality of transmission estimation in optical networks

Disaggregated data centers (DCs) can offer high flexibility for resource allocation; hence, their resource utilization can be significantly improved. However, communications between different types of resources, in particular between CPU and memory, in fully disaggregated DCs face severe problems in terms of stringent requirements for ultra-low latency and ultra-high transmission bandwidth. Optical fiber communication is promising to provide high capacity and low latency, but it is still challenging for the state-of-the-art optical technologies to meet the requirements of fully disaggregated DCs. In this article, different levels of resource disaggregation are investigated. For fully disaggregated DCs, two architectural options are presented, where optical interconnects are necessary for the CPU-memory communications. We review the state-of-the-art optical transmission and switching technologies, and analyze pros and cons of their applicability in the disaggregated DCs. The results reveal that resource disaggregation does improve the resource utilization in DCs. However, the bandwidth provided by the state-of-the-art technologies is not always sufficient for fully disaggregated DCs. It calls for further advances in optical communications to fully utilize the advantages of fully disaggregated DCs.

/pdf/disaggregated-data-centers-challenges-and-trade-offs-2p4y21f8k9.pdf

Disaggregated Data Centers: Challenges and Trade-offs

Digital-to-Analog Converters (DACs) are a key technology for high-speed optical links and are widely deployed. A new and improved model for a DAC that accurately accounts for the frequency dependent nature of ENoB and is computationally efficient is validated through both simulations and experiments. Furthermore, the impact of the various model parameters such as quantization, timing jitter and bandwidth, on the performance of 16QAM, 64QAM, and 256QAM are presented.

Frequency Dependent ENoB Requirements for M-QAM Optical Links: An Analysis Using an Improved Digital to Analog Converter Model

Short reach optical interconnects must support higher data rates to manage the increasing needs of end users and the commensurate increase in storage and computation within and between data centers. VCSEL-based MMF links provide a power efficient solution which is achieved in part by maximizing the data rate per laser source. We design an archetype system to explore the challenges and benefits of signal processing with currently available technology. By careful use of signal processing at both the transmitter and receiver, we experimentally demonstrate 107 Gb/s PAM-4 transmission over 105 m of wideband MMF using 850 and 940 nm VCSELs and thus show a path to 400 Gb/s and eventually 1 Tb/s multimode fiber links.

DSP-Enabled 100 Gb/s PAM-4 VCSEL MMF Links

We demonstrate a path to 1-Tb/s VCSEL-based links using a wideband multimode fiber (WBMMF/OM5), which provides an extended effective modal bandwidth ranging from 844 to 948 nm. Using 850-, 880-, 910-, and 940-nm vertical cavity surface emitting lasers (VCSELs), we quantify the benefits of electronic pulse-shaping and channel equalization for both pulse-amplitude modulations (PAMs), i.e., PAM-2 and PAM-4. We demonstrate that all these VCSELs support an error-free performance (bit error rate <10−12) at 54 Gb/s for PAM-2 and at 84 Gb/s for PAM-4 over 100 m of OM5 without forward error-correction (FEC) or receiver equalization. Furthermore, all VCSELs support 100-Gb/s PAM-4 transmission over 100 m of OM5 with FEC and receiver equalization. Importantly, all signal filtering investigated is readily implemented in SiCMOS.

Scaling VCSEL-MMF Links to 1 Tb/s Using Short Wavelength Division Multiplexing

We demonstrate one-class SVM employing readily available receiver side DSP features to detect soft link failures. The technique was able to detect inferior lasers, faulty ROADMs, OSNR degradation and inter-channel interference with <4% classification error.

Identification of Soft Failures in Optical Links using Low Complexity Anomaly Detection

We experimentally demonstrate accurate modulation format identification, optical signal to noise ratio (OSNR) estimation, and bit error ratio (BER) estimation of optical signals for wavelength division multiplexed optical communication systems using convolutional neural networks (CNN). We assess the benefits and challenges of extracting information at two distinct points within the demodulation process: immediately after timing recovery and immediately prior to symbol unmapping. For the former, we use 3D Stokes-space based signal representations. For the latter, we use conventional I-Q constellation images created using demodulated symbols. We demonstrate these methods on simulated and experimental dual-polarized waveforms for 32-GBaud QPSK, 8QAM, 16QAM, and 32QAM. Our results show that CNN extracts distinct and learnable features at both the early stage of demodulation where the information can be used to optimize subsequent stages and near the end of demodulation where the constellation images are readily available. Modulation format identification is demonstrated with >99.8% accuracy, OSNR estimation with <0.5 dB average discrepancy and BER estimation with percentage error of <25%.

Siddharth Varughese

Papers

Frequency Dependent ENoB Requirements for M-QAM Optical Links: An Analysis Using an Improved Digital to Analog Converter Model

DSP-Enabled 100 Gb/s PAM-4 VCSEL MMF Links

Scaling VCSEL-MMF Links to 1 Tb/s Using Short Wavelength Division Multiplexing

Identification of Soft Failures in Optical Links using Low Complexity Anomaly Detection

Optical performance monitoring using digital coherent receivers and convolutional neural networks