Integrated Photonic Neural Networks for Equalizing Optical Communication Signals: A Review
The demand for high-capacity communication systems has grown exponentially in recent decades, constituting a technological field in constant change. Data transmission at high rates, reaching tens of Gb/s, and over distances that can reach hundreds of kilometers, still faces barriers to improvement, such as distortions in the transmitted signals. Such distortions include chromatic dispersion, which causes a broadening of the transmitted pulse. Therefore, the development of solutions for the adequate recovery of such signals distorted by the complex dynamics of the transmission channel currently constitutes an open problem since, despite the existence of well-known and efficient equalization techniques, these have limitations in terms of processing time, hardware complexity, and especially energy consumption. In this scenario, this paper discusses the emergence of photonic neural networks as a promising alternative for equalizing optical communication signals. Thus, this review focuses on the applications, challenges, and opportunities of implementing integrated photonic neural networks for the scenario of optical signal equalization. The main work carried out, ongoing investigations, and possibilities for new research directions are also addressed. From this review, it can be concluded that perceptron photonic neural networks perform slightly better in equalizing signals transmitted over greater distances than reservoir computing photonic neural networks, but with signals at lower data rates. It is important to emphasize that photonics research has been growing exponentially in recent years, so it is beyond the scope of this review to address all existing applications of integrated photonic neural networks.
Understanding Fabrication Variability in Core-Shell Soft Biomaterials Using Stochastic Artificial Intelligence
Approaches for the fabrication of biomaterials are currently numerous, with a wide diversity of available material precursors, chemistries, and processing technologies. Owing to the complex nature of the human body, biomaterials are targeted for applications with highly diverse performance demands. Traditional strategies based on trial and error have fallen short of predicting the ideal parameters required to produce adequate structures to meet these challenges. Although the design of experiments enables reducing experimental testing, it has failed to predict complex, multi-factorial processing effects, including experimental variability. Despite being often overlooked, experimental variability is an important aspect in biomaterials, which are often processed from source materials with significant compositional variability (e.g., natural polymers), along with processing methodologies frequently undertaken under poorly controlled environmental conditions. Here, a machine learning approach based on Gaussian processes (GPs) is developed to identify patterns and correlations between fabrication conditions and material properties. Flexible soft membrane-based tubular materials obtained by polyelectrolyte complexation are used as a model biomaterial characterized by multi-parametric design inputs. Using GPs, the effects of processing parameters on the magnitude and variability of key properties like permeability, porosity, thickness, opacity, and swelling ratio are quantified. This approach is expected to enable more reliable and predictable biomaterial fabrication.
Network Design in Multiple Non-Overlapping Paths within Quantum Key Distribution Networks
The increasing demand for secure communication makes Quantum Key Distribution (QKD) a compelling technology, as it exploits the fundamental principles of quantum physics to enhance secure data transmission. However, practical deployment faces limitations such as limited transmission distances and dedicated hardware-channel requirements. This paper proposes a linear programming-based model that jointly addresses routing, wavelength assignment, and trusted-node minimization in QKD networks with wavelength division multiplexing (WDM), considering that the secret key rate becomes impractical beyond a given transmission distance, thus requiring the relaying of secret keys through trusted repeaters. In particular, the study contrasts scenarios where traffic is concentrated along a single route with those where key rates are split and routed over multiple nonoverlapping paths (MNOPs) using two mechanisms: Concatenate Multipart (CM) and Bitwise Product Multipath (BPM). While the literature shows that MNOPs can improve security (BPM) and improve the secret key rate (CM), we demonstrate that these mechanisms come with additional deployment costs.
Photonic Computing Towards Energy-Efficient AI Solutions Based on Photoactivated Materials
Artificial intelligence (AI) and neuromorphic computing are driving the development of hardware platforms that combine high energy efficiency with physically informed data processing. Photonic technologies offer distinct advantages, particularly when coupled with photoactivated materials that allow light-matter interactions to be directly exploited for computation. Inspired by synaptic mechanisms in biological systems, a neuromorphic photonic layer based on a luminescent material was developed to mimic synaptic memory via the material’s optical afterglow. The information-processing capability of the photonic front-end was evaluated by coupling its output responses to a digital machine learning model, achieving 76.9% test accuracy on the MNIST handwritten digit recognition task. This approach demonstrates the potential of hardware AI systems that tightly integrate material-level functionalities with neuromorphic computation by exploiting synapse-like optical memory in luminescent materials.
Highly Stretchable Eu3+ – Doped PEBAX Elastomers With Complementary Optical Readouts of Strain and Temperature
Soft photonic materials capable of maintaining optical functionality under large mechanical deformation are highly desirable for wearable photonics, flexible sensors, optical and mechanically adaptive photonic systems. Here, we report highly stretchable luminescent elastomers based on the thermoplastic copolymer PEBAX 2533 incorporating a europium β-diketonate complex. The transparent films combine intense europium emission with exceptional mechanical flexibility, displaying elongation at break exceeding 1200%, while maintaining high optical transparency in the visible range. The elastomers exhibit intense red luminescence with absolute quantum yields of 0.61. Under mechanical deformation, the 5D0 → 7F0 transition shows a strain-dependent spectral shift and intensity modulation, enabling optical strain readout with sensitivities of –1.6 ± 0.2 cm−1·ε−1 and −0.44 ± 0.04 ε−1, respectively. The luminescence spectra also display a thermal dependence, characterized by a blue shift and intensity decrease of the Eu3+ emission with a temperature sensitivity of 0.08 ± 0.01 cm−1·°C−1 over the investigated temperature range. The coexistence of large deformability, high transparency, efficient Eu3+ emission, and two independently calibrated optical responses establishes Eu-doped elastomers as a versatile soft photonic platform for multimodal optical transduction, highlighting their potential for flexible sensing technologies in which mechanical and thermal perturbations are optically encoded within the same material platform.
Post consumer expanded polystyrene as an optical-grade host for lanthanide-doped luminescent solar concentrators
The transition toward net-zero buildings has increased interest in building-integrated photovoltaics (BIPV). Luminescent solar concentrators (LSCs) are particularly attractive due to their transparency, colour tunability, and seamless architectural integration. Yet most LSCs rely on virgin polymers, including polystyrene (PS), a fossil-derived polymer. Expanded polystyrene (EPS) waste is difficult to recycle because of degradation, contamination, and low material value. This study addresses this challenge by converting post-consumer EPS into an optical-grade matrix for LSCs, giving a second life to a hard-to-recycle plastic. Recycled PS (r-PS) was doped with two lanthanide complexes — Eu(TFNB)3Phen and Tb(4BBA)3Phen — and analysed structurally and photophysically. The r-PS retained high transparency and achieved quantum yields of 0.53 ± 0.05 (Eu) and 0.40 ± 0.04 (Tb). Its mechanical properties were preserved, enabling fabrication of functional LSCs with power conversion and optical efficiencies of (0.013 ± 0.001)% and (1.2 ± 0.1)%, respectively. The devices also exhibited dual-mode thermometric sensitivity up to 3.9 ± 0.6 % °C−1. A life-cycle assessment showed significant reductions in acidification, terrestrial eutrophication, and resource use when virgin PS is replaced with r-PS.
Security-aware non-uniform multipath traffic partitioning for QKD networks: A MILP-based squeezing protection approach
Quantum Key Distribution (QKD) enables information-theoretic secure key exchange, but the deployment of large-scale QKD networks introduces new challenges related to routing, resource allocation, and security exposure in trusted-node-based architectures. In multipath QKD networks, the secret key rate can be partitioned and transmitted over multiple link-disjoint paths, so that an adversary must simultaneously compromise all paths to reconstruct the complete secret key. However, traffic is often equally distributed across the available disjoint paths, although such uniform partitioning may be suboptimal when candidate paths exhibit heterogeneous vulnerability levels. This paper proposes a security-aware optimization framework for QKD-enabled optical networks that determines a non-uniform multipath traffic partitioning strategy aimed at minimizing the expected amount of eavesdropped traffic. The proposed approach is formulated as a Mixed-Integer Linear Programming (MILP) model that integrates probabilistic secrecy metrics derived from trusted node compromise probabilities, a traffic vulnerability threshold that limits partial information exposure, and adaptive bandwidth squeezing protection to ensure service continuity under failure scenarios. The model jointly optimizes traffic allocation and excess key capacity, allowing the protection overhead to be dynamically determined according to security and resilience requirements. Numerical evaluations conducted on both a toy topology and the Pan-European network demonstrate that the proposed exposure-aware partitioning strategy significantly reduces the expected eavesdropped traffic compared to conventional equal-splitting approaches while maintaining resilience guarantees. These results highlight the potential of security-driven traffic engineering for the design of resilient and efficient multipath QKD networks.
The Role of Photonics in Neuromorphic Computing for Augmenting the von Neumann Paradigm
The rapid expansion of Big Data and Artificial Intelligence has exposed fundamental inefficiencies in conventional von Neumann computing architectures, thus necessitating a search for alternative solutions. Neuromorphic computing, inspired by biological neural systems, has emerged as a promising route toward faster and more energy-efficient information processing approaches. Optical neuromorphic computing exploits light as a signal modality for stimulation, data delivery, information processing, and signal readout, offering intrinsic advantages in speed, parallelism, and power consumption. This review outlines the core principles of neuromorphic computing and examines recent advances in its optical implementations, with a particular focus on optical artificial synapses, photonic layers integrated into hybrid photonic-digital architectures, and optical neural networks. These emerging approaches define a coherent framework for advancing computing systems that offer clear advantages over purely von Neumann systems.
Mobile-compatible neuromorphic optical computing enabled by dual-emission photonic materials
Artificial intelligence (AI) and neuromorphic computing demand hardware platforms that combine energy efficiency with physically informed data processing. Photonics offers unique advantages in this context, but optical neuromorphic systems in which memory and information processing arise from intrinsic material dynamics remain scarce. We demonstrate a photonic layer in which history-dependent material photophysics implements physically embedded information processing before digital learning, using luminescent phosphors. The photonic layer exhibits dual fluorescence and phosphorescence together with excitation-history-dependent photoactivation dynamics that emulate synaptic functionalities, including short-term memory, long-term memory, and synaptic potentiation. Quantitative analysis and modelling reveal efficient nanoscale interlamellar energy transfer consistent with the lamellar material morphology, establishing a link between structure and function. When integrated as an active optical front-end within a hybrid photonic–digital AI architecture, the photonic layer performs a material-based transformation of input data before digital learning. Using a laboratory-based optical readout, classification accuracy comparable to state-of-the-art hybrid neuromorphic computing (approximately 94%) is achieved while requiring fewer training epochs. A mobile-compatible implementation based on smartphone optical readout yields similar accuracy, demonstrating robustness to non-specialized optical hardware. These results outline a general strategy for neuromorphic photonic systems in which functional materials act as adaptive computational primitives, enabling energy-efficient computing architectures.
Tunable multicomponent emission from carbon nanomaterials for neuromorphic optical synaptic functions
Multi-component luminescent materials provide an effective route to engineer complex optical responses and temporal emission dynamics for neuromorphic photonics. In this work, we present a neuromorphic photonic layer based on a multi-component carbon-based nanomaterial (THR@BA) composed of a rigid boron oxide matrix derived from the calcination of boric acid (BA), which embeds three organic molecules, terephthalic acid (TPA), 1-hydroxy-2-naphthoic acid (HNA) and rhodamine B (RhB), forming a confined host-guest system with enhanced intermolecular interactions. The rigid matrix promotes energy transfer and excited-state coupling between the molecular components. The multi-component nanomaterial exhibits dual emission of fluorescence and long-lived phosphorescence and simultaneously provides complex responses upon multi-wavelength excitation. These responses result from the superposition of the optical properties of each single-component nanomaterial functionalized with the respective organic molecule, namely TPA@BA, HNA@BA, and RhB@BA. By tuning the excitation parameters, THR@BA reproduces key synaptic functions, including short- and long-term memory as well as potentiation involved in synaptic plasticity, deriving from the dynamic balance between fluorescence and phosphorescence emission with lifetimes ranging from nanoseconds to seconds. This work demonstrates that multi-component molecular systems confined within simple inorganic matrices can serve as versatile platforms for engineering programmable luminescence dynamics, making them ideal candidates for artificial optical synapses and neuromorphic computing.