Search

High-speed, low-power optical interconnects, such as intensity modulation direct detection (IMDD) optical links, are increasingly deployed in data centers to keep pace with the growing bandwidth requirements. High-sensitivity low-power optical receivers (RXs) are the key components that enable energy-efficient IMDD optical interconnects. This article presents a low-power nonreturn-to-zero (NRZ) optical RX using a combination of a limited-bandwidth trans-impedance amplifier (TIA) and duobinary sampling to improve RX sensitivity at high data rates. Duobinary sampling leverages the well-controlled TIA inter-symbol interference (ISI) to recover the transmitted data, making it much more hardware efficient than canceling the ISI using a decision feedback equalizer (DFE). The proposed optical RX employs a CMOS-based analog front-end (AFE) to achieve high linearity and excellent power efficiency. Fabricated in 65-nm CMOS process, the prototype RX achieves optical modulation amplitude (OMA) sensitivity of −11.6 dBm at 16 Gb/s with 0.7-pJ/bit efficiency.

An optoelectronic millimeter-wave (mm-wave) radar transmitter (TX) integrated circuit (IC) for multiple-input–multiple-output (MIMO) imaging radar in silicon photonics technology is presented. The MIMO radar system architecture in which the ICs shall be used features an optical local oscillator (LO) distributed over fiber to coherently operate mm-wave transceiver frontends. Optical LO distribution enables larger apertures and hence higher angular resolution for MIMO imaging radars. The optoelectronic TX frontend IC integrates all electronic and photonic components needed to receive the optical LO signal and to generate a transmit signal at mm-wave. It operates on both 1310- and 1550-nm LO wavelengths. The chip was designed in a pre-production photonic SiGe BiCMOS technology from IHP. Measured output power is 13 dBm at 67 GHz and 5 dBm at 77 GHz from −5-dBm optical modulation amplitude (OMA). The IC dissipates 900 mW from a single 3.6-V supply voltage.

In this article, we demonstrate a push-pull silicon dual-ring modulator based on two cascaded rings which are coupled with each other through two parallel bus waveguides. In the transmission spectrum of the modulator, a narrow transparency peak exists in the center of the broad dip based on the electromagnetically induced transparency (EIT)-like effect. Benefiting from both the EIT-like effect and push-pull drive configuration, a low insertion loss (IL) of 1.4 dB and an extinction ratio (ER) over 6.4 dB are obtained at a bit rate of 10 Gb/s with a peak-to-peak driving voltage (Vpp) of 5 V. The corresponding maximum optical modulation amplitude (OMA) is 0.56, which is almost twice that of a single micro-ring modulator with the same driving voltage. Moreover, the maximum OMA is 0.47 at 20 Gb/s, which is 81% larger than that of a single micro-ring modulator.

Multi-channel optical receivers clocked at a frequency slower than the data-rate in which different phases are used for each channel are widely used in high data-rate applications. Here, we demonstrate a two-channel optical receiver where both channels share the same clock operating at half of the data-rate frequency. However, instead of using different phases for each channel, the optical input is split and the input to the 2nd channel is optically delayed with respect to the 1st channel. Moreover, to improve performance, the duty cycle of the clock can easily become a design parameter. Each channel consists of only one high-bandwidth gain-improved transimpedance amplifier with a pseudo-differential-output and a comparator with offset-nulling. The receiver is fabricated in 65 nm CMOS. To demonstrate the concept, the fabricated die and two photo-detectors are bonded inside a QFN80 package. Measurement results of the receiver show that each channel can sample the input at 8.5 Gbps, resulting in a 17 Gbps total data rate, with a total energy efficiency of 156 fJ/bit, and an input optical modulation amplitude (OMA) sensitivity of −7 dBm without any equalization.

For the silicon photonic links to meet the target bit error rate (BER), the laser source must output enough optical power to overcome the optical channel loss and limited receiver sensitivity. Combined with poor wall-plug efficiency, this optical power requirement makes the electrical power consumed by the laser source a significant portion of the link energy cost. This article proposes a laser-forwarded coherent link that greatly reduces the required laser optical power by averaging optical path loss between the main and forwarded paths and by improving the receiver sensitivity using a homodyne coherent detector. The link performance analysis reveals that the laser forwarded link improves the laser power budget by ≈6–8 dB compared with a conventional intensity-modulation/direct detection link using typical photonic link components. To demonstrate the laser forwarded link, a microring resonator phase modulator, a balanced detector, and a 3-dB coupler are integrated with CMOS circuits in a GFUS 45-nm SOI process. The transmit driver and the receiver consume 40 and 450 fJ/bit, respectively. Aided by ≈8-dB boost from laser forwarding and the coherent detection gain, the receiver achieves −15.6-dBm optical modulation amplitude sensitivity. The link operates at 10 Gb/s with BER < 10−9 and electrical energy efficiency of 2.3 pJ/bit.

This article presents an electro-absorption modulator (EAM)-based single-mode (SM) 50-Gb/s non return to zero (NRZ) Si-photonic optical link in a 16-nm FinFET. The EAM device is modeled to include its electrical and optical properties. The transmitter (TX) uses T-coil based over-peaking to improve modulation efficiency and relax transimpedance amplifier’s (TIA’s) bandwidth and noise requirement. The receiver (RX) uses a three-stage TIA with T-coils to improve bandwidth. The link sensitivity is −10.9-dBm optical modulation amplitude (OMA) at bit error rate (BER) < 10−12 with 2-dB link margin at 50 Gb/s. The combined power efficiency of the RX, TX, and clocking is 3.16 pJ/bit and the external laser consumes 1.15 pJ/bit with 10% wall-plug efficiency.

A fully integrated 25 Gb/s low-noise optical receiver is presented which integrates transimpedance amplifier (TIA), continuous-time linear equalizer (CTLE), high-gain and high-bandwidth limiting amplifier (LA), and clock and data recovery (CDR) circuit into a single die. The TIA employs an inverter-based pseudo-differential TIA scheme with input series-inductor peaking, cross-coupled negative Gm pair and negative capacitance to improve the bandwidth, and noise performance, while a MOSFET corner compensation (MCC) circuit compensates for CMOS corner variations. A gain control (GC) scheme is proposed which solves the group delay issue caused by TIA input impedance variations from small input to overload current. Finally, a 2x-oversampling CDR using a bang-bang phase detector is included. The receiver is fabricated in 40-nm CMOS process, and the 850 nm VCSEL-based full-link measurement results show that the optical receiver achieves 44 $\mu \text{A}_{\mathrm {pp}}$ (RSSI Current = 43 $\mu \text{A}$ , ER = 4.94 dB) optical modulation amplitude (OMA) sensitivity (BER

A 25Gb/s guardring-free planar AlInAs avalanche photodiode (APD) and 100-Gb/s APD receiver optical sub-assembly (ROSA) employing a quasi-collimated beam for reducing the optical loss and optical demultiplexer (DEMUX) size are demonstrated for a 40-km optical link. The APD so fabricated achieves a very low dark current of 63 nA at a multiplication factor of 10, with a high gain-bandwidth product of 225 GHz due to the optimum multiplication layer and absorption layer. Additionally, the assembled APD ROSA exhibits a high receiver sensitivity of more than −23 dBm optical modulation amplitude (OMA) (BER = 5.0 × 10−5) at temperatures ranging from −5 to 80 °C for each lane. Investigating the transmission over fibers with dispersions of −119 and +40 ps/nm reveals that a sufficiently low penalty of less than 0.9 dB is obtained for all lanes. Furthermore, it is confirmed that the APD ROSA suffers little crosstalk penalty thanks to the isolation achieved by this design of optical DEMUX.

This letter presents a 3-D-integrated 26 Gb/s opto-electrical receiver front-end. The electronic integrated circuit (EIC) is fabricated in a BiCMOS-55-nm technology, flipped and placed on top of the photonic integrated circuits (PICs) die through copper pillars. In the receiver chain, a fully differential shunt-feedback TI amplifier (FD-SF TIA) is followed by a limiting amplifiers (LAs) with embedded equalization, output driver and an automatic offset cancelation loop. The whole receiver provides a transimpedance (TI) gain of 76 dB $\Omega $ with 30-GHz bandwidth. By exploiting the FD-SF TIA with low parasitic capacitance of the Germanium dual heterojunction photo diode (Ge-PD) in the photonic die, the receiver achieves sensitivity of −15.2 dBm optical modulation amplitude (OMA) at Ge-PD and −10-dBm OMA at the single-mode fiber (SMF) optical output with bit error rate of ${10^{ - 12}}$ and PRBS 15. The sensitivity is aligned with state-of-the-art receivers employing discrete photonics and, to author’s best knowledge, it is the lowest reported among published 25 Gb/s receivers exploiting silicon photonics.

This paper describes an avalanche photodetector (APD)-based optical receiver, applicable to the burst-mode operation, in 28-nm CMOS technology. With the aims of benefiting the overall optical link power efficiency and link bandwidth, the optical receiver is designed to have high sensitivity and high reconfiguration speed for burst-mode operation. The sensitivity of the receiver is optimized by adjusting the responsivity of APD via its reverse bias voltage, which leads to the highest signal-to-noise ratio (SNR) at the front end. The two-tap feed-forward equalization (FFE), along with two-tap decision feedback equalization, is implemented in a current-integrating fashion to further improve the sensitivity with superior power efficiency to their resistively loaded counterparts. Integrating dc comparator and integrating amplitude comparator are proposed to replace the conventional RC low-pass filters and peak detectors, respectively, in extracting the information of dc offset and signal amplitude within two unit intervals (UIs), empowering significant acceleration of the burst-mode reconfiguration. When the APD is biased at −16 V, its overall responsivity at 1310 nm is 4 A/W, and the optical receiver achieves bit-error-rate (BER) better than 10−12 at −16-dBm optical modulation amplitude, 2.24-ns reconfiguration time with 5-dB dynamic range, and 1.37-pJ/b energy efficiency at 25 Gb/s.

In this, we propose a novel design of a silicon photonic modulator that has a high modulation efficiency and that is tolerant to temperature variations. A series of phase-shifted Bragg gratings are placed in each arm of a Mach–Zehnder interferometer in order to provide enhanced phase modulation. The slow light effect in these ultra-compact coupled resonators improves phase modulation efficiency compared to conventional silicon phase shifters. These Bragg grating cavities are designed such that the optical bandwidth is increased compared to other coupled resonators, such as micro-rings. This improved bandwidth reduces the temperature sensitivity of the devices. We present in detail how to optimize these modulators considering properties, such as modulation efficiency ( Vπ × L ), optical modulation amplitude, and optical bandwidth ( $\Delta \lambda _{BW}$ ); the latter property determines the operating temperature range (Δ T ). As examples, we present two designs that meet different target specifications for short-reach or long-haul applications. We further provide a model, based on the coupled mode theory, to investigate the dynamic response of the proposed modulators. A large signal analysis is performed using finite-difference time domain in order to simulate on/off keying modulation and eye diagrams up to 110 Gb/s.

We present the design and implementation of a 90 -Gb/s non-return-to-zero (NRZ) direct detection optical receiver that consists of a low-noise transimpedance amplifier (TIA), fabricated in a 55-nm SiGe BiCMOS technology, and a Ge photodiode integrated into a Silicon Photonic integrated circuit. Low-noise broadband operation is achieved using a fully differential TIA that provides the photodiode reverse bias. 50-, 56-, and 64-Gb/s NRZ operation is demonstrated at bit-error ratios (BERs) below $10^{-12}$ at sensitivities of $-$ 12.3-, $-$ 11.2-, and $-$ 11.1-dBm optical modulation amplitude (OMA) while consuming 128, 141, and 157 mW, respectively. Furthermore, a BER below $10^{-12}$ is achieved at 80 Gb/s with a sensitivity of $-$ 6.1 dBm (OMA) at 165 mW, and operation below KP4-FEC $({2.4 \times {10}^{-4}})$ is demonstrated up to 90 Gb/s with a sensitivity of $-$ 7.1 dBm (OMA) while consuming 222 mW. All BERs were measured in real time and without DSP or equalization to compensate the receiver. The presented receiver achieves the best reported sensitivities up to 90 Gb/s, at the second-lowest power consumption, and is the first reported integrated optical RX up to 90 Gb/s that is tested in real time without any equalization or DSP to compensate the receiver. The presented optical receiver is ideally suited for upcoming short-reach applications.

This paper demonstrates a receiver (RX) for optical communications implemented in a 28-nm digital bulk-CMOS technology. Compact bandwidth (BW)-enhancement methods, such as inductor sharing and stacking, result in a measured BW of 27 GHz. By using these area-efficient techniques and a mixed-signal offset compensation system, a small active area of 0.009 mm2 is achieved. Noise and jitter are analyzed for the transimpedance amplifier and the limiting amplifier. Operation up to a data rate (DR) of 54 Gbit/s is demonstrated electrically with an energy efficiency of 0.6 pJ/bit. A commercial, 25-Gbit/s photo-diode (PD) is wire-bonded to the RX chip on a printed circuit board (PCB) assembly. Error-free (BER < 10−12) optical transmission at 53 Gbit/s is achieved with a transmitter (TX) using a 35-GHz electro-optical modulator. At this DR, an energy-per-bit of 0.65 pJ/bit is measured without the 16.5-mW output buffer, which only serves measurement purposes. An input sensitivity of −6-dBm optical modulation amplitude (OMA) is measured at 53 Gbit/s. Power/DR adaptivity is featured by the design; so when the DR can be reduced, the power consumption is decreased in order to maintain the energy efficiency. Energy-per-bit is improved from 1.28 pJ/bit by 52% to 0.67 pJ/bit at 27 Gbit/s.

We present a novel optical phase modulation scheme on a Si photonic platform that uses a III–V/Si hybrid metal–oxide–semiconductor (MOS) capacitor formed by bonding an n-type InGaAsP membrane on a p-type Si waveguide. We numerically revealed that the phase modulation efficiency was improved by a factor of 7–8 owing to electron accumulation at the InGaAsP MOS interface when the n-type Si layer in a Si MOS optical phase shifter was replaced by an n-type InGaAsP layer. To realize the III–V/Si hybrid MOS capacitor, we developed an Al2O3 bonding interface deposited by atomic layer deposition that enabled a low interface trap density of μ m wavelength owing to the electron-induced change in the refractive index of InGaAsP. Since no holes were induced in the III–V layer of the III–V/Si hybrid MOS capacitor, we avoided large hole-induced absorption in InGaAsP. As a result, when we had a π phase shift, we obtained optical absorption of 0.23 dB, approximately ten times smaller than that of a Si MOS optical phase shifter. We found by numerical analysis that the efficient low-loss III–V/Si hybrid MOS optical phase shifter improved markedly the optical modulation amplitude, indicating its suitability for high-speed modulation beyond 100 Gb/s. We also demonstrated a Mach–Zehnder interferometer optical switch using the proposed optical phase shifter with a switching time of less than 20 ns. We achieved an extremely low switching power of approximately 1 nW, enabling a large-scale optical switch and universal photonic integrated circuits. We also discuss the feasibility of a photonic neural network for deep learning.

This book covers the technology of switching or modulating light in semiconductor optical waveguides. Currently a key function for optical communications systems is the conversion of data from an electrical signal to an optical signal for transmission in very low loss optical fibres and the converse process of optical to electrical conversion the O/E/O data conversion. It focuses on all-optical switching using second/third order optical nonlinearities in AIGaAs optical waveguides while covering a variety of device configurations including integrated nonlinear couplers and Mach-Zehnder interferometers. Supplementary Mathematica programs are available within Book Information.

In this work, we demonstrate In0.52Al0.48As top/backside-illuminated avalanche photodiodes (APD) with dual multiplication layers for high-speed and wide dynamic range performances. Our fabricated top-illuminated APDs, with a partially depleted p-type In0.53Ga0.47As absorber layer and thin In0.52Al0.48As dual multiplication (M-) layer (60 and 88 nm), exhibit a wide optical-to-electrical bandwidth (16 GHz) with high responsivity (2.5 A/W) under strong light illumination (around 1 mW). The measured bias dependent 3-dB O-E bandwidth was pinned at 16 GHz without any serious degradation near the saturation current output. To further increase the speed, we downscaled the active diameter and adopted a back-side illuminated structure with flip-chip bonding for batter optical alignment tolerance. A significant improvement in maximum bandwidth was demonstrated (25 versus 18 GHz). On the other hand, we adopted a thick dual M-layer (200 and 300 nm) and 2 μm absorber layer in the APD design to circumvent the problem of serious bandwidth degradation under high gain (&gt, 100) and high-power operation which significantly enhanced the dynamic range. Due to dual M-layer, the carriers could be energized in the first M-layer then propagate to the second M-layer to trigger the avalanche process. In both cases, despite variation in thickness of the absorber and M-layer, the cascade avalanche process leads to values close to the ultra-high gain bandwidth product (GBP) of around 460 GHz with a responsivity of 0.4 and 1 A/W at unit gain for the thin and thick M-layer devices, respectively. We successfully achieved a good sensitivity of around −20.6 dBm optical modulation amplitude (OMA) at a data rate of 25.78 Gb/s, by packaging the fabricated APDs (thin dual M-layer (60 and 88 nm) version) with a 25 Gb/s trans-impedance amplifier in a 100 Gb/s ROSA package. The results show that, the incorporation of a dual multiplication (M) layer structure in the APD opens a new window to obtaining the higher GBP in order to meet the requirements for high-speed transmission without the need of further downscaling the multiplication layer.

25-Gbit/s non-return-to-zero (NRZ) transmissions were successfully extended to 100 km by using a 1358-nm-wavelength SOA integrated EADFB laser (AXEL) with an optical modulation amplitude (OMA) over 10.5 dBm.

We numerically analyzed the modulation characteristics of an InP organic hybrid (IOH) optical modulator consisting of an InP slot waveguide and an electro-optic (EO) polymer. Since InP has a higher electron mobility and a lower electron-induced free-carrier absorption than Si, the series resistance of an InP slot waveguide can be significantly reduced with relatively smaller optical loss than an Si slot waveguide. As a result, the trade-off between optical loss and modulation bandwidth can be remarkably improved compared with a Si organic hybrid (SOH) optical modulator. When the modulation bandwidth was designed to be 100 GHz, the optical loss of the IOH modulator was 13-fold smaller than that of the SOH one. The simulation of the eye diagram revealed that the improved optical modulation amplitude enabled the clear eye opening with a 100 Gbps non return-to-zero signal using the IOH modulator. The IOH integration is promising for a high-speed modulator with low energy consumption beyond 100 Gbps.

In this paper, a simulation link is established with proposed electric-optical modulator models. Optical modulation amplitude (OMA) and bit error rate (BER) of these modulators are studied in combination with a simulation link with a transmission rate of 60 Gbps. In the simulation, the total lengths of a single arm of these modulators is 3750 um, and the peak-to-peak voltage (VPP) of each input electrical signal is controlled at 2 V, and the input optical power of the laser is set at 20 mW. The results show that with the increasing level of pulse amplitude modulation (PAM), the higher BER appears while OMA is insensitive to the change of level of pulse amplitude modulation. And when PAM-16 is implemented, relatively clear eye diagram is still obtained. It proves that the application prospect of this modulator is in higher level pulse amplitude modulation, such as PAM-8 and PAM-16. Therefore, SEMZM is a good choice for high speed data transmission.

We present a custom-designed integrated circuit (IC) implemented in 0.25- μ m BiCMOS technology that can automatically control the Si micro-ring modulator (MRM) temperature for optimal modulation characteristics. The IC monitors the optical modulation amplitude (OMA) of an Si MRM and provides the optimal heater setting for the maximum OMA. The IC consists of trans-impedance amplifier, power detector, track-and-hold circuit, comparator, digital-to-analog converter, and digital controller, all of which are integrated in a single chip. We demonstrate that, with this IC, an Si MRM can provide the maximum OMA for 25-Gb/s operation despite changes in temperature and input optical power.

Doping with foreign atoms is an efficient way to enhance the electrochromic performance of the pristine single material. In this work, N-doped TiO2-B nanotube films were successfully prepared via hydrothermal approach combined with ion exchange method. The nitrogen doped TiO2-B nanotube film shows good electrochromic performance and cyclic stability. At 633 nm and −1.6 V, the optical modulation amplitude of the nitrogen doped TiO2-B film can reach 76%, which is higher than pristine TiO2-B film (64%). The coloring efficiency of the nitrogen doped samples is 35.38 cm2/C which is higher than pure TiO2-B (21.72 cm2/C). Such good performance can be ascribed to the increase of the TiO2-B electrical conductivity caused by the doping of N species.

This brief presents a single-ended asymmetric push-pull inverter for driving a vertical-cavity surface-emitting laser (VCSEL). The proposed driver topology features less power dissipation compared with commonly used differential CML-based drivers. Considering power efficiency and the bandwidth of the VCSEL, the modulation current and the bias of the VCSEL are set to 1.5 and 6.2 mA, respectively. Diode-connected cascode transistors are employed to ensure the adequate voltage level at the output node. Series inductive peaking is utilized to extend the limited bandwidth by splitting the load capacitance and the driver self-capacitance. The proposed VCSEL driver has been fabricated in 65-nm CMOS technology and occupies an active area of 0.009 mm2. The extinction ratio and the optical modulation amplitude of the driver are measured to be 1.8 dB and 0 dBm at 35 Gb/s. The chip consumes 22.6 mW at the data rate of 35 Gb/s, which corresponds to the energy efficiency of 0.65 pJ/b.

High-speed top-illuminated avalanche photodiodes (APDs) with large diameters (25 μ m) are demonstrated for the application of 4-channels 100 Gb/s data rate. They achieve a bandwidth of 17 GHz at low-gain (MG = 6.2; 3.6 A/W) and large-gain bandwidth (responsivity bandwidth) product (410 GHz (237.8 GHz-A/W); 55% external efficiency at the unit gain) while maintaining invariant high speed (14 GHz) under high power (0.5 mW) and 0.9 Vbr operations. By packaging the demonstrated APD with a 25 Gb/s transimpedance amplifier in a 100 Gb/s ROSA package, a good sensitivity of around –20.6 dBm optical modulation amplitude (OMA) at the data rate of 25.78 Gb/s has been successfully demonstrated. The achieved sensitivity not only meets the required receiver sensitivity (–18.5 dBm OMA) in 100 GbE-ER-4 Lite (40 km) system, it is also comparable with that of the high-performance 100 Gb/s ROSA incorporated with the back-side illuminated APD design. Overall, our novel APD structure can eliminate the costly flip-chip bonding package in the 100 Gb/s ROSA without sacrificing its sensitivity performance.

Silicon photonic (SiP) electro-optic modulators are a key component in cost-efficient and integrated optical transmitters. Modulator design traditionally uses figure of merits (FOMs) that characterize modulation efficiency and propagation loss of light, which underestimate the modulator-induced power penalty due to intersymbol interference, as they do not consider the electro-optic bandwidth limitation. We show that in the presence of limited electro-optic bandwidth of the SiP modulator, the conventional FOMs, such as VπL and V παL , are unable to predict the minimum transmitter power penalty (TPP). Normalized optical modulation amplitude (OMAN) is proved through simulation to be a reliable tool to predict the minimal TPP point. Then, we introduce a new FOM that includes not only the efficiency of the modulator, but also the bandwidth limitation from the SiP electro-optic modulator. The new FOM that is derived from OMAN translates the system-level requirements of a PAM- M optical link to the device-level design parameters. This FOM can be hired to optimize driving voltage swing, bias voltage, and phase-shifter length or to simply choose a SiP modulator with minimal imposed TPP.

This paper reports an interesting phenomenon of optical modulation enhancement produced by CW optical sideband injection locking of a directly modulated Fabry-Perot Laser Diode (FPLD) lasing at 1557.1 nm wavelength, the modulation being sinusoidal in nature and effected at 300 kHz and 3 MHz. This investigation is carried out for direct modulation of the FPLD effected in the RF region at 300 kHz and 3 MHz in contradistinction to earlier studies being conducted with the modulation in the microwave frequency band. A nonlinear analysis has been carried out which corroborates experimental observation on modulation enhancement. At low level of optical Intensity Modulation (IM) index, not exceeding 20 %, typical modulation enhancement of 2 % at 300 kHz and 11 % at 3 MHz modulation frequency have been observed. Larger modulation enhancement is expected at higher values of IM indices of the slave FPLD and for medium to high power laser diodes. This circuit can have direct application in Wavelength Division Multiplexed-Passive Optical Network (WDM-PON).

A universal analytical design approach for laser diode drivers (LDDs) is presented and verified. All design parameters are derived analytically, taking the vertical-cavity surface-emitting laser (VCSEL) characteristics into account. Two optimization strategies are proposed. First, the power consumption is minimized for the highest achievable optical modulation amplitude (OMA) and extinction ratio (ER). Second, the highest achievable data rate (DR) is considered to maximize the energy efficiency. As a result, a simple differential amplifier can be implemented as an LDD to achieve energy efficiencies and DRs comparable or even higher than the one of more complex designs utilizing preemphasis, feedforward equalization, and pulse amplitude modulation. Therefore, drivers designed with this holistic analytical approach accommodate well the demands for high DRs, high energy efficiency, high compactness, high reliability, and low latency in future optical data links. A compact LDD is designed with an active area of only $0.15 \times 0.12$ mm2 in a 130-nm SiGe BiCMOS technology. A DR of 40 Gbit/s with bit error rates $ and an energy efficiency of 2.245 pJ/bit are achieved for an OMA of 1.1 mW and an ER of 10 dB using a 20-GHz 850-nm common-cathode VCSEL. The achieved values deviate less than 10% from the predicted ones, which clearly illustrates the effectiveness of the presented analytical approach. Further improvements are achieved by optimizing the modulation levels and the performance to the link requirements. At 45 and 30 Gbit/s, 1.8 and 1.17 pJ/bit are achieved, respectively.

Utilizing four-level pulse-amplitude modulation (PAM-4) with vertical-cavity surface emitting lasers (VCSELs) is challenging due to device nonlinearity and bandwidth limitations. This letter presents a VCSEL-based PAM-4 transmitter that employs a 2.5-tap nonlinear equalizer to compensate for these issues. The transmitter utilizes lookup-table control of a current-mode digital-to-analog converter (DAC) output stage to enable independent equalization coefficients for different PAM-4 symbol transitions. Utilizing half pre-cursor (MSB only), one main, and one post-cursor taps provides a good balance between circuit complexity and performance. Equalization resolution is increased by segmenting the output stage into a 2-bit symbol DAC and a fine 3-bit equalization DAC. Fabricated in GP 65-nm CMOS, the serializing transmitter operates at 50 Gb/s with 3-dBm optical modulation amplitude (OMA) when driving an 11-GHz VCSEL and achieves 5.12 pJ/bit power efficiency.

A 50-Gb/s optical receiver (RX) based on 0.18- $\mu \text{m}$ SiGe BiCMOS technology was fabricated and evaluated. To improve the horizontal eye opening and sensitivity of the RX without degrading its power efficiency, it is configured with a two-stage pre-amplifier with high gain (56 dB $\Omega $ ) and high bandwidth (35 GHz), a high-accuracy automatic decision-threshold control (ATC) consisting of power-supply variation and offset cancellers for improving sensitivity, and a jitter-reduction packaging structure for improving transimpedance flatness. The RX achieves high sensitivity of −9.2 dBm, while its power consumption is 7.9 pJ/bit at a data rate of 50 Gb/s. In 50- and 56-Gb/s transmission tests through a 100-m multi-mode fiber (MMF), it demonstrated error-free operation at the bit error rate (BER) of less than 10–12 with the sensitivities of −9 and −6.1 dBm optical modulation amplitude (OMA). It also demonstrated 300-m error-free MMF transmission (at 50 Gb/s) with the practical sensitivity of −7.6 dBm OMA.

In this paper, the morphological and optical properties of LTO/TaOx composite structure films and electrochromic properties of “ITO/NiOx/LTO/TaOx/WOx/ITO” electrochromic devices (ECDs) based on LTO/TaOx composite structure films are reported. The composite films are deposited on the ITO substrate by reactive magnetron sputtering. The results of XRD, SEM, and AFM show that LTO film is amorphous and the performance of films prepared by sputtering power of 100 W for 5 h is the best. The optical transmittance of the multilayer structure is tested using an ultraviolet-visible spectrum and the refractive index dispersion is calculated using the relationship between the peaks in the spectrum based on equal inclination interference theory. Optical transmittance and refractive index dispersion are changed with the increase of magnetron sputtering power. Electrochromic tests show that all-solid-state ECDs based on composite films have 13 s coloring and 14 s fading response time, and the average optical modulation amplitude is 58%.

This paper presents a 32 Gb/s non-return-to-zero optical link using 850-nm vertical-cavity surface-emitting laser-based multi-mode optics with 14-nm bulk FinFET CMOS circuits. The target application is the integration of optics on to the first-level package, connecting high-speed optical I/O directly to an advanced CMOS host chip (e.g., processor and switch) to increase package I/O bandwidth density and lower overall system power and cost. The optical link is designed for maximum link margin to tolerate high optical losses created by low-cost optical packaging. The transmitter (TX) uses a three-tap, 1/2-unit-interval-spaced feed-forward equalizer to improve eye opening. The receiver (RX) uses a low-bandwidth, low-noise transimpedance amplifier and a speculative one-tap decision-feedback equalizer for high sensitivity. The TX and RX power efficiencies are 3.3 and 1.4 pJ/bit, respectively. The TX optical modulation amplitude (OMA) is 1.2 dBm, and the RX sensitivity is −11.7 dBm OMA at a bit error rate of 10−12 with PRBS31 data, providing 12.9-dB link margin.

We report on a simple way to generate complex optical waveforms with very cheap and accessible equipments. The general idea consists in modulating a laser diode with an autonomous electronic oscillator, and in the case of this study, we use a distributed feedback (DFB) laser diode pumped with an electronic Chua’s circuit. Based on the adiabatic P-I characteristics of the laser diode at low frequencies, we show that when the total pump is greater than the laser threshold, it is possible to convert the electrical waveforms of the Chua’s circuit into optical carriers. But, if that is not the case, the on-off dynamical behavior of the laser permits to obtain many other optical waveform signals, mainly pulses. Our numerical results are consistent with experimental measurements. The work presents the advantage of extending the range of possible chaotic dynamics of the laser diodes in the time domains (millisecond) where it is not usually expected with conventional modulation techniques. Moreover, this new technique of laser diodes modulation brings a general benefit in the physical equipment, reduces their cost and congestion so that, it can constitute a step towards photonic integrated circuits.

We experimentally demonstrate cost effective symmetrical 112-Gbps long-reach passive optical network (LR-PON) over 70-km standard signal mode fiber (SSMF), based on pulse amplitude modulation (PAM)-4. Four 10G-class directly modulated lasers (DMLs) at C-band are used for achieving 4 × 28 -Gbps downstream transmission, while two 18G-class DMLs at O-band are used to realize 2 × 56 -Gbps upstream transmission, without any optical amplification in optical distributed network (ODN). Both dispersion compensation fiber (DCF) for downstream signal and praseodymium-doped fiber amplifier (PDFA) for upstream signal are equipped at optical line terminal (OLT). Meanwhile, sparse Volterra filter (SVF) equalizer is proposed to mitigate the transmission impairments with substantial reduction of computation complexity. Finally, we can successfully provide a loss budget of 33 dB per downstream wavelength channel, indicating of 64 optical network units (ONUs) with more than 1.25 Gbps per ONU.

Optical interconnects are being increasingly deployed in data centers to meet growing bandwidth requirements under tight power constraints. Therefore, there has been renewed focus on increasing data rates and improving power efficiency of optical links. Among all the link components, laser diodes typically consume the most power. Their power dissipation is dictated by the amount of signal power that needs to be transmitted to meet the bit error rate requirements under given channel loss and receiver sensitivity conditions. Consequently, improving receiver sensitivity directly helps lowering laser diode power consumption. In this paper, we present design techniques to implement such high sensitivity optical receivers. To this end, we identify noise–bandwidth tradeoffs of a shunt feedback transimpedance amplifier (SF-TIA) and elucidate how they limit the maximum achievable sensitivity at a given data rate and process technology. We then propose to combine a low-bandwidth SF-TIA with a four-tap decision feedback equalizer to overcome this noise–bandwidth tradeoff. The proposed SF-TIA uses high-gain multistage amplifier and large feedback resistance and achieves greatly improved noise performance. Fabricated in a 65-nm CMOS technology and heterogeneously integrated with a photonic IC, the proposed optical receiver achieves optical modulation amplitude sensitivity of −16.8 dBm with 1.9-pJ/bit efficiency at 12 Gb/s.

A scheme for reconfigurable channel slicing and stitching is proposed and experimentally demonstrated. By employing optical nonlinear wave mixing and a coherent frequency comb, a single channel spectrum is sliced and redistributed into fragmented frequency slots, which can be stitched together to recover the original channel at the receiver. This approach is verified through a single channel experiment with the modulation formats of quadrature phase-shift keying and 16 quadrature amplitude modulation. The system exhibits less than 1.5% error-vector-magnitude deterioration and no more than 2-dB optical signal-to-noise ratio penalty, compared to a back-to-back baseline. To demonstrate robustness of the scheme, different parameters of the channel slices are varied, such as relative phase offset, relative amplitude, and the number of slices. A 10-km transmission experiment is also conducted and the additional system penalty is negligible. This scheme is used to experimentally demonstrate fragmented channel bandwidth allocation in a dense 6-channel wavelength-division-multiplexing system. The incoming 20-Gbaud optical channel is successfully reallocated into two fragmented frequency slots and reconstructed at the receiver.

In this letter, the design and measurement results of a high-swing 42-Gb/s vertical-cavity surface-emitting laser (VCSEL) driver with an optical modulation amplitude of 1.4 dB is presented. To keep the power dissipation low, no equalizer was employed at the highest rate. At the output stage circuit, a series-shunt peaking technique is used, and it is optimized to lower the group-delay variation at the laser input. The transmitter circuitry can be disabled when there is no data transmission to save half of the power required for the full-rate functionality. Then, it can be enabled in less than 10 ns to full rate operation. The chip is fabricated in a 14-nm bulk CMOS technology and bonded to a common-cathode 20-GHz VCSEL. Optical measurements show that the data transmission up to a data rate of 42 Gb/s with bit error rate better than 10−12 was possible. The total power dissipation, including the one of the VCSEL, is 81.5 mW, which provides a power efficiency of 1.94 pJ/b. To the best of our knowledge, this is the fastest VCSEL driver in any CMOS process suitable for burst mode operation.

This paper presents recent results in the design, fabrication, and demonstration of indium phosphide based integrated optoelectronic devices for applications in communications and signal processing. The underlying theme in the design and use of these devices is to exploit the parallelism that manifests itself owing to the availability of a set of stabilized, mutually coherent optical carriers, or optical frequency combs.

We demonstrate an integrated circuit (IC) implemented in 28-nm CMOS technology for Mach-Zehnder modulator (MZM) bias control. Our circuit determines the optimal MZM bias voltage that provides the largest optical modulation amplitude and maintains this bias voltage by monitoring the average modulated power. The controller IC consists of digital-to-analog converter, analog-to-digital converter, trans-impedance amplifier, power detector, track-and-hold circuit, comparator, and digital controller, all of which are integrated in a single chip. With small power consumption and chip area, our approach should be applicable for future photonic electronic integrated circuits based on the Si platform.

The next generations of large-scale data-centers and supercomputers demand optical interconnects to migrate to 400G and beyond. Microring modulators in silicon-photonics VLSI chips are promising devices to meet this demand due to their energy efficiency and compatibility with dense wavelength division multiplexed chip-to-chip optical I/O. Higher order pulse amplitude modulation (PAM) schemes can be exploited to mitigate their fundamental energy–bandwidth tradeoff at the system level for high data rates. In this paper, we propose an optical digital-to-analog converter based on a segmented microring resonator, capable of operating at 20 GS/s with improved linearity over conventional optical multi-level generators that can be used in a variety of applications such as optical arbitrary waveform generators and PAM transmitters. Using this technique, we demonstrate a PAM-4 transmitter that directly converts the digital data into optical levels in a commercially available 45-nm SOI CMOS process. We achieved 40-Gb/s PAM-4 transmission at 42-fJ/b modulator and driver energies, and 685-fJ/b total transmitter energy efficiency with an area bandwidth density of 0.67 Tb/s/mm2. The transmitter incorporates a thermal tuning feedback loop to address the thermal and process variations of microrings’ resonance wavelength. This scheme is suitable for system-on-chip applications with a large number of I/O links, such as switches and general-purpose and specialized processors in large-scale computing and storage systems.

We investigate and compare the performance of star-shaped 16-ary quadrature amplitude modulation (Star-16QAM) and square-shaped 16QAM (Square-16QAM) in the probabilistic shaping (PS) and uniform schemes with coherent detection. With the help of PS technology, the bit error ratio (BER) improvement achieved in the PS-Star-16QAM scheme is greater than that of the PS-Square-16QAM when compared with the uniform schemes in our numerical simulation and experiment. Therefore, the PS-Star-16QAM shows superiority over the PS-Square-16QAM in terms of the BER improvement.

By combining optical Fourier transform with second-harmonic generation (SHG) from infrared to visible region in a periodically poled Magnesium-doped Lithium Niobate (PPLN) crystal, we present a simple yet efficient experiment to revisit the optical implementation of image self-convolution. In contrast to previous scheme, here the main feature is the holographic gratings we prepare by a phase-only spatial light modulator that enables a deeper understanding on the centrosymmetry for intensity images, and particularly, the conjugate symmetry for both intensity and phase images. We manifest both cases by observing the real-time autocorrelation peaks that appear gradually in the output SHG signals. As our optical scheme possesses the inherent capacity of parallel processing, it will find potential in pattern recognition, particularly, for infrared remote sensing.

Since energy dissipation and latency in optical interconnects are of utmost concern, such links are often operated without forward error correction. We propose a low-complexity noniterative two-error correcting Bose–Chaudhuri–Hocquenghem (BCH) decoder circuit that significantly relaxes the stringent optical modulation amplitude requirements on the transmitter, which help to reduce laser and laser driver energy dissipation. We demonstrate, in a 28-nm CMOS process technology, that the introduction of this BCH circuit in an uncoded link leads to energy-per-bit reductions of 25%, even when the impact of code rate on receiver energy dissipation is considered.

We have developed a dual-output, DFB-LD array type, widely wavelength tunable laser in the L-band. The device can serve simultaneously as the signal light source of the transmission section and the local light emission source of the reception section in the digital coherent transmission system. By using dual-output TLD array, compact module is realized, avoiding occurrence of problems such as insertion loss generation, increase in power consumption, as compared with a configuration in which an optical coupler for output branching is connected to one high-power TLD module. The chip power consumption was 0.58 W at 27 °C, and 0.91 W at 55 °C at two fiber powers of 13.0 and 10.0 dBm, respectively. Estimated module power consumption was 3.0 W or less. These characteristics are suitable for digital coherent transmission of 100 Gbps.

In a self-mixing interferometer based on a diode laser, the measurement of the external reflector displacement $s(t)$ is carried out by looking at the optical-phase signal $\cos 2ks(t)$ , a signal readily detected as an amplitude modulation (AM) of emitted power. In contrast, the other available signal, sin $2ks(t)$ , a frequency modulation (FM) of the emitted field at optical frequency, is never used because difficult to recover. Recently, Contreras et al. used a narrow-band acetylene cell to convert the FM into an amplitude signal, finding it is larger and has a better SNR than the AM. In this paper, we analyze the advantages of the new CFM (converted-FM) signal, calculating both amplitude and SNR, and compare theoretical results to published experimental evidence, finding good agreement. We then present options for realizing the selective filter in different configurations and technology. Finally, we evaluate the improvement offered by CFM in a number of measurements, like sub-wavelength vibrations, digital readout displacement, and diode laser alpha factor.

The systematic design analysis of a hybrid optoelectronic limiting receiver is presented. The limiting receiver was designed in the IBM 130-nm CMOS process and side-by-side wire bonded to a commercial high-speed InGaAs/InP PIN photodiode. The CMOS die and the optical die were chip-on-board (COB) mounted on the Printed Circuit Board. Partition between small signal in the linear region and large-signal in the limiting region is emphasized. The signal edges can get sharpened in the limiting region although the front-end circuit is bandwidth-limited. When tested with a PRBS-31 pattern, the prototype achieved a bit-error rate of $10^{-12}$ at the sensitivity level of −3.2-dBm Mach-Zehnder Modulator optical modulation amplitude at 4 Gb/s. Conclusions derived from this paper can provide insights to guide other optical limiting receiver design and testing at higher speeds.

We extend the transmission distance for 28-Gb/s optical links from 40 to 80 km in a single-mode fiber (SMF) in the 1.3-μm-wavelength window by using an intensity-modulation direct-detection scheme. To extend the link budget, and compensate for the optical loss of SMF, we try to increase the optical output power of an electroabsorption modulator integrated with distributed feedback (EADFB) laser. We apply a semiconductor optical amplifier (SOA)-assisted extended-reach EADFB laser (AXEL) to a 1.3-μm-wavelength light source. To increase the optical output power while using energy efficiently, we design the SOA length in terms of achieving a balance between the laser diode (LD) and the SOA length with a constant injection current of 160 mA. We fabricate an AXEL and an EADFB laser and compare their basic characteristics. By using the high-power AXEL, we achieve the first 28-Gb/s 80-km SMF transmission without the need for digital signal processing or an in-line optical amplifier. The AXEL consumes only 0.26 W and does not require any digital processing or an in-line amplifier and can, therefore, reduce the power needed for an optical network system by tens of watts.

A novel scheme which can realize high-speed bidirectional optical signals transmission and reception by using 4-pulsed amplitude modulation (4-PAM) and demodulation in 10 Gb/s passive optical access systems, is introduced in this paper. In this scheme, a PAM sequence generator module, a M-ary pulse generator module and one Mach-Zehnder intensity Modulator (MZM) are employed for generating the 10 Gb/s optical 4-PAM signals. The generated optical 4-PAM signal is achieved full-duplex transmission over 20 km single mode fiber (SMF) for both downstream (DS) and upstream (US) access links application. The electrical spectrum diagrams, optical spectrum diagrams, and eye-diagrams are obtained, and the receiver sensitivities of 10 Gb/s 4-PAM signals before and after bidirectional transmission are analyzed. The simulation results demonstrate that, while the 4-PAM signal is adopted in passive optical access systems as bidirectional transmission signal, it can be received effectively with the BER below the forward-error-correction (FEC) limit value.