arXiv subjects

Emiliano Staffoli

Publications and source records attributed to Emiliano Staffoli.

6 recordsLinked to original sources

arXiv2026

Memory in Integrated Photonic Neural Networks: From Physical Mechanisms to Neuromorphic Architectures

The rapid scaling of artificial neural networks has exposed fundamental limitations of conventional von Neumann computing architectures. In these systems, the physical separation between memory and processing creates a bottleneck, as computational capabilities outpace the ability of memory and interconnects to supply and retrieve data. In contrast, biological neural systems inherently co-localize computation and memory through distributed, dynamical processes. Neuromorphic computing seeks to emulate this paradigm by leveraging physical substrates whose intrinsic dynamics simultaneously encode and process information. Among emerging platforms, silicon photoncis offer a compelling approach due to its high bandwidth, low-loss propagation, and inherent parallelism. This review examines the role of memory in integrated photonic neuromorphic systems, with emphasis on the physical mechanisms that provide volatile (short-term) and non-volatile (long-term) memory in silicon-on-insulator and hybrid silicon-on-insulator platforms. Drawing inspiration from digital, biological, and photonic memory architectures, we classify existing approaches based on their underlying physical principles. We cover implementations ranging from delay lines and slow-light structures to multistable dynamics and structural memory based on charge trapping and phase-change materials. We then discuss how these mechanisms support photonic neural network architectures, including feed-forward, reservoir computing, spiking and hybrid optoelectronic recurrent systems, and assess their relevance for time-dependent singal-processing tasks such as channel equalization in telecommunications. This review aims to establish a unified framework for understanding memory and learning in neuromorphic photonics and outlines key challenges and opportunities for scalable, energy-efficient neuromorphic hardware.

physics.optics↗

arXiv2025

Nonlinear Distortion Equalization in Multi-Span Optical Links Via a Feed-Forward Photonic Neural Network

Linear and nonlinear distortions in optical communication signals are equalized using an integrated feed-forward Photonic Neural Network (PNN). The PNN is based on a linear stage made of an 8-tap Finite Impulse Response (FIR) filter, featuring tunable amplitude and phase weights at each tap, and of a nonlinear stage achieved through the square modulus operation at the end-of-line photodetector. Within an Intensity Modulation/Direct Detection (IMDD) system, the PNN is applied to 2-level Pulse Amplitude Modulated (PAM2) optical signals undergoing multi-span propagation. Each 50 km segment includes fiber transmission, optical power restoration, and optional chromatic dispersion compensation via a Tunable Dispersion Compensator. Positioned at the receiver, the PNN enables fully optical signal processing with minimal latency and power consumption. Experimental validation is conducted using a Silicon-On-Insulator device operating on 10 Gbps signals. It demonstrates chromatic dispersion equalization over distances up to 200 km and self-phase modulation (with dispersion removed) up to 450 km. Simulations explore PNN adaptation for 100 Gbps modulations and its potential for cross-phase modulation equalization.

physics.optics↗

arXiv2025

A Silicon Photonic Neural Network for Chromatic Dispersion Compensation in 20 Gbps PAM4 Signal at 125 km and Its Scalability up to 100 Gbps

A feed-forward photonic neural network (PNN) is tested for chromatic dispersion compensation in Intensity Modulation/Direct Detection optical links. The PNN is based on a sequence of linear and nonlinear transformations. The linear stage is constituted by an 8-tap time-delayed complex perceptron implemented on a Silicon-On-insulator platform and acting as a tunable optical filter. The nonlinear stage is provided by the square modulus of the electrical field applied at the end-of-line photodetector. The training maximizes the separation between the optical levels (i.e. the eye diagram aperture), with consequent reduction of the Bit Error Rate. Effective equalization is experimentally demonstrated for 20 Gbps 4-level Pulse Amplitude Modulated signal up to 125 km. An evolutionary algorithm and a gradient-based approach are tested for the training and then compared in terms of repeatability and convergence time. The optimal weights resulting from the training are interpreted in light of the theoretical transfer function of the optical fiber. Finally, a simulative study proves the scalability of the layout to larger bandwidths, up to 100 Gbps.

cs.ET↗

arXiv2023

Photonic neural networks based on integrated silicon microresonators

The recent progress of artificial intelligence (AI) has boosted the computational possibilities in fields where standard computers are not able to perform. The AI paradigm is to emulate human intelligence and therefore breaks the familiar architecture on which digital computers are based. In particular, neuromorphic computing, artificial neural networks (ANN) and deep learning models mimic how the brain computes. Large networks of interconnected neurons whose synapsis are individually strengthened or weakened during the learning phase find many applications. With this respect, photonics is a suitable platform to implement ANN hardware thanks to its speed, low power dissipation and multi-wavelength opportunities. One photonic device candidate to perform as an optical neuron is the optical microring resonator. Indeed microring resonators show both a nonlinear response and a capability of optical energy storage, which can be interpreted as a fading memory. Moreover, by using silicon photonics, the photonic integrated circuits can be fabricated in volume and with integrated electronics on board. For these reasons, here, we describe the physics of silicon microring resonators and of arrays of microring resonators for application in neuromorphic computing. We describe different types of ANNs from feed-forward networks to photonics extreme learning machines and reservoir computing. In addition, we discuss also hybrid systems where silicon microresonators are coupled to other active materials. this review aims to introduce the basics and to discuss the most recent developments in the field.

physics.optics↗

arXiv2023

Equalization of a 10 Gbps IMDD signal by a small silicon photonics time delayed neural network

A small 4-channels time-delayed complex perceptron is used as a silicon photonics neural network (NN) device to compensate for chromatic dispersion in optical fiber links. The NN device is experimentally tested with non-return-to-zero optical signals at 10 Gbps after propagation through up to 125 km optical fiber link. During the learning phase, a separation-loss function is optimized in order to maximally separate the transmitted levels of 0s from the 1s, which implies an optimization of the bit-error-rate. Testing of the NN device shows that the excess losses introduced by the NN device are compensated by the gain in transmitted signal equalization for a link longer than 100 km. The measured data are reproduced by a model which accounts for the optical link and the neural network device. This allows simulating the network performances for higher data rates, where the device shows improvement with respect to the benchmark both in terms of performance as well as ease of use.

cs.ET↗

arXiv2023

Nonlinear response of Silicon Photonics microresonators for reservoir computing neural network

Nowadays, Information Photonics is extensively studied and sees applications in many fields. The interest in this breakthrough technology is mainly stimulated by the possibility of achieving real-time data processing for high-bandwidth applications, still implemented through small-footprint devices that would allow for breaking the limit imposed by Moore's law. One potential breakthrough implementation of information photonics is via integrated photonic circuits. Within this approach, the most suitable computational scheme is achieved by integrated photonic neural networks. In this chapter, we provide a review of one possible way to implement a neural network by using silicon photonics. Specifically, we review the work we performed at the Nanoscience Laboratory of the University of Trento. We present methodologies, results, and future challenges about a delayed complex perceptron for fast data processing, a microring resonator exploiting nonlinear dynamics for a reservoir computing approach, and a microring resonator with the addition of a feedback delay loop for time series processing.

cs.ET↗