1,721,059 research outputs found
InP Monolithically Integrated Transmitters Based on High Speed Directly Modulated DFB Lasers
No full textIn this paper we summarize results concerning high-speed integrated multi-wavelength transmitters fabricated using a generic InP integration platform. The photonic integrated circuits (PICs) include tunable directly modulated DFB lasers emitting in the C-band, monitoring photodetectors, MMI couplers and output spot size converters. The directly modulated lasers, designed in order to fit within a 100 GHz grid, show a maximum 3-dB small signal bandwidth of 21 GHz and a tunability range of about 4 nm. In particular, we report an 8-channel transmitter showing operation up to 30 Gb/s per channel (for 200+ Gb/s applications) and a 4-channel transmitter capable of operating up to 34 Gb/s per channel (for 100+ Gb/s applications). Error free operation (bit error rates <10-9) is achieved for all the channels in all the cases
A Lithium Niobate on Insulator Based Photonic Neural Network
No full textWe propose and simulate a photonic neuromorphic engine that exploits the Lithium-Niobate-on-Insulator (LNOI) integration platform. The resulting reduced-precision vector-matrix multiplier can trade off speed with resolution at a reduced power consumption enabled by LNOI platform
Photonic Neural Networks: A Survey
No full textPhotonic solutions are today a mature industrial reality concerning high speed, high throughput data communication and switching infrastructures. It is still a matter of investigation to what extent photonics will play a role in next-generation computing architectures. In particular, due to the recent outstanding achievements of artificial neural networks, there is a big interest in trying to improve their speed and energy efficiency by exploiting photonic-based hardware instead of electronic-based hardware. In this work we review the state-of-the-art of photonic artificial neural networks. We propose a taxonomy of the existing solutions (categorized into multilayer perceptrons, convolutional neural networks, spiking neural networks, and reservoir computing) with emphasis on proof-of-concept implementations. We also survey the specific approaches developed for training photonic neural networks. Finally we discuss the open challenges and highlight the most promising future research directions in this field
Activation Stretching for Tackling Noise in Photonic Aware Neural Networks
No full textThis paper introduces a stretching strategy for nonlinear activation functions aimed to enhance noise resilience in photonic-aware neural networks. Its effectiveness is numerically demonstrated in counteracting different noise levels in low-resolution operations. (c) 2023 The Author(s
Lithium Niobate on Insulator Technology for applications in the 1064 nm Window
No full textThin film lithium niobate (TFLN) enables a high-performance integration platform thanks to its superior electro-optic properties and broad transparency range. In this contribution, we discuss preliminary results toward the exploitation of TFLN in the 1064nm window, further enhancing its versatility for telecommunications and sensing applications
An overview of hardware acceleration techniques for 5G functions
No full textNetwork Function Virtualisation (NFV) is emerging as a promising solution for lowering capital expenditures and ease network function upgrade. However, typical virtualised CPUs cannot provide the processing performance required by some network functions. Thus, some type of optimized acceleration is needed. In this paper, an overview of OpenCL-based hardware acceleration techniques for 5G functions, specifically IFFT and Cyclic Prefix Addition Process, is provided. Implementations and possible optimization are discussed
Resolution Limits in Optical Microring Modulators for Applications in Neuromorphic Photonics
No full textNeuromorphic photonic devices are promising for overcoming limitations in AI processing faced by traditional computing architectures. An analytical technique for determining the resolution limits introduced by photonic link distortions and noise is applied to microring modulators, with optimal resolution exhibited by a critically coupled ring
Comb Line Multiplication in an Integrated Optical Frequency Comb Generator
No full textWe demonstrate frequency-comb tone multiplication in an InP photonic integrated comb generator made of cascaded modulators. Tone multiplication is achieved driving the dual-drive MZM on the chip with an RF signal and one submultiple
- …
