Realization Of Networks Research Articles

Optimizing telemetry forwarding for distributed failure recovery in packet-optical networks

Fast network recoverability from hard and soft failures is crucial for network operators to deliver uninterrupted services. Streaming telemetry has been studied as a solution for enabling fast and accurate failure detection in optical networks. However, significant delay is incurred when relying on a centralized entity (e.g., software-defined network controller) to collect, process, and act on telemetry data. Programmable switches (e.g., P4-based) allow telemetry data to be processed at line speed, enabling local on-device (distributed) decisions. These devices can be used to deploy quick and local mitigation to failures while a global solution is being computed on a longer time scale. However, designing network-wide streaming telemetry with distributed decisions remains an open challenge. In this work, we specify the joint optimization of packet-optical networks with on-device failure recovery, considering multiple aspects of the problem. The problem is modeled using linear programming and solved for multiple network realizations. The solutions can be used to program each switch in the network to detect failures and quickly recover the traffic. Results show that the proposed model decreases the required number of register entries to store telemetry data while assuring high recoverability and a minimized number of wavelengths.

Compositional Sparsity, Approximation Classes, and Parametric Transport Equations

Abstract Approximating functions of a large number of variables poses particular challenges often subsumed under the term “Curse of Dimensionality” (CoD). Unless the approximated function exhibits a very high level of smoothness the CoD can be avoided only by exploiting some typically hidden structural sparsity. In this paper we propose a general framework for new model classes of functions in high dimensions. They are based on suitable notions of compositional dimension-sparsity quantifying, on a continuous level, approximability by compositions with certain structural properties. In particular, this describes scenarios where deep neural networks can avoid the CoD. The relevance of these concepts is demonstrated for solution manifolds of parametric transport equations. For such PDEs parameter-to-solution maps do not enjoy the type of high order regularity that helps to avoid the CoD by more conventional methods in other model scenarios. Compositional sparsity is shown to serve as the key mechanism for proving that sparsity of problem data is inherited in a quantifiable way by the solution manifold. In particular, one obtains convergence rates for deep neural network realizations showing that the CoD is indeed avoided.

Physical Realizations of Interdependent Networks: Analogy to Percolation

Percolation on interdependent networks generalizes the well-studied percolation model in a single network to multiple interacting systems, unveiling spontaneous cascading failures, abrupt collapses, and high vulnerability. The main novelty of interdependent networks has been the introduction of two types of links, connectivity within networks and the dependency between them. The interplay between these two types of interactions results in novel critical phenomena and phase transitions. This abstract percolation paradigm was recently applied to magnetic networks, as an experimentally testable method for interdependent superconducting networks as well as to other systems like k-core percolation and overloaded networks. Here, we will review these physical applications and provide insights into several potential directions for the field of physically interdependent networks.

CDUNeXt: efficient ossification segmentation with large kernel and dual cross gate attention

Ossification of the ligamentum flavum (OLF) is the main causative factor of spinal stenosis, but how to accurately and efficiently identify the ossification region is a clinical pain point and an urgent problem to be solved. Currently, we can only rely on the doctor’s subjective experience for identification, with low efficiency and large error. In this study, a deep learning method is introduced for the first time into the diagnosis of ligamentum flavum ossificans, we proposed a lightweight, automatic and efficient method for identifying ossified regions, called CDUNeXt. By designing lightweight module structures, utilizing large-kernel convolutions to extracts the long-distance dependencies of different features of the image, and adopting dual-cross-gate-attention(DCGA) to sequentially capture the channel and spatial dependencies so as to fast and accurate segmentation while maintaining fewer parameters and lower complexity. Experiments show that CDUNeXt achieves the best segmentation performance with an optimal balance of lighter weights and less computational cost compared to existing methods. This work fills the gap in the application of deep learning techniques in the diagnosis of ligamentum flavum ossificans, contributes to the realization of lightweight medical image segmentation networks and lays the foundation for subsequent research.

Delayed-feedback oscillators replicate the dynamics of multiplex networks: Wavefront propagation and stochastic resonance

The widespread development and use of neural networks have significantly enriched a wide range of computer algorithms and promise higher speed at lower cost. However, the imitation of neural networks by means of modern computing substrates is highly inefficient, whereas physical realization of large scale networks remains challenging. Fortunately, delayed-feedback oscillators, being much easier to realize experimentally, represent promising candidates for the empirical implementation of neural networks and next generation computing architectures. In the current research, we demonstrate that coupled bistable delayed-feedback oscillators emulate a multilayer network, where one single-layer network is connected to another single-layer network through coupling between replica nodes, i.e. the multiplex network. We show that all the aspects of the multiplexing impact on wavefront propagation and stochastic resonance identified in multilayer networks of bistable oscillators are entirely reproduced in the dynamics of time-delay oscillators. In particular, varying the coupling strength allows suppressing and enhancing the effect of stochastic resonance, as well as controlling the speed and direction of both deterministic and stochastic wavefront propagation. All the considered effects are studied in numerical simulations and confirmed in physical experiments, showing an excellent correspondence and disclosing thereby the robustness of the observed phenomena.

Fast delivery of heralded atom-photon quantum correlation over 12 km fiber through multiplexing enhancement

Distributing quantum entanglement between distant parties is a significant but difficult task in quantum information science, as it can enable numerous applications but suffers from exponential decay in the quantum channel. Quantum repeaters are one of the most promising approaches towards this goal. In a quantum repeater protocol, it is essential that the entanglement generation speed within each elementary link is faster than the memory decoherence rate, and this stringent requirement has not been implemented over a fiber of metropolitan scale so far. As a step towards this challenging goal, in this work we experimentally realize multiplexing-enhanced generation of heralded atom-photon quantum correlation over a 12 km fiber. We successively generate 280 pairs of atom-photon quantum correlations with a train of photonic time-bin pulses filling the long fiber, and read out the excited memory modes on demand with either fixed or variable storage time after successful heralding. With the multiplexing enhancement, the heralding rate of atom-photon correlation can reach 1.95 kHz, and the ratio between the quantum correlation generation rate to memory decoherence rate can be improved to 0.46 for a fiber length of 12 km. This work therefore constitutes an important step towards the realization of a large-scale quantum repeater network.

Water-to-air unmanned aerial vehicle (UAV) based rolling shutter optical camera communication (OCC) system with gated recurrent unit neural network (GRU-NN).

Underwater sensors and autonomous underwater vehicles (AUVs) are widely adopted in oceanic research activities. As the number of underwater sensors and AUVs is growing quickly, the bandwidth requirements are increasing accordingly. In this work, we put forward and demonstrate a large field-of-view (FOV) water-to-air unmanned aerial vehicle (UAV) based optical camera communication (OCC) system with gated recurrent unit neural network (GRU-NN) for the first time to the best of our knowledge. As the UAVs are embedded with complementary-metal-oxide-semiconductor (CMOS) cameras, there is no need to install OCC receivers (Rxs), reducing the deployment cost. Moreover, the large photo-sensitive area of the CMOS camera can support a large FOV OCC transmission without the need for precise optical alignment. Here, by utilizing the column matrix identification during the rolling shutter pattern decoding in the CMOS image sensor, the scintillation caused by water turbulence can be reduced. Besides, in the outdoor and windy environment, the UAV will experience significant movement caused by the wind making it very difficult to capture stable OCC frames in the CMOS image sensor. Here, we propose and demonstrate utilizing GRU-NN, which is a special realization of the recurrent neural network (RNN) with memory cells capable of learning the time-domain dependent signals. It is shown that the GRU-NN can learn effectively from successive image frames in time-domain and produce correct prediction even under the windy and unstable UAV flying environment. Experimental results reveal that the proposed GRU-NN can outperform the previous pixel-per-symbol labeling neural network (PPS-NN), and also can significantly reduce the computation time when compared with long-short-term-memory-neural-network (LSTM-NN). The proposed system can decode 4-level pulse-amplitude-modulation (PAM4) rolling shutter OCC patterns at data rates of 5.4 kbit/s and 3.0 kbit/s under clear and cloudy water, respectively, fulfilling the pre-forward error correction bit-error rate (pre-FEC BER = 3.8 × 10-3). We also demonstrate that the UAV based OCC system can support data rates of 5.4 kbit/s, 4.2 kbit/s, and 3.0 kbit/s at distances of 2.2 m, 3.2 m, and 4.2 m, respectively, at outdoor and windy environments.

Enforcing exact permutation and rotational symmetries in the application of quantum neural networks on point cloud datasets

Recent developments in the field of quantum machine learning have promoted the idea of incorporating physical symmetries in the structure of quantum circuits. A crucial milestone in this area is the realization of Sn-permutation equivariant quantum neural networks (QNNs) that are equivariant under permutations of input objects. In this paper, we focus on encoding the rotational symmetry of point cloud datasets into the QNN. The key insight of the approach is that all rotationally invariant functions with vector inputs are equivalent to a function with inputs of vector inner products. We provide a structure of the QNN that is exactly invariant to both rotations and permutations, with its efficacy demonstrated numerically in the problems of two-dimensional image classifications and identifying high-energy particle decays, produced by proton-proton collisions, with the SO(1,3) Lorentz symmetry. Published by the American Physical Society 2024

Optical activation function using a metamaterial waveguide for an all-optical neural network.

In this study, we experimentally demonstrated that the nonlinear optical coefficient of the original Si can be enhanced by incorporating a metamaterial structure into an existing silicon waveguide. The two-photon absorption coefficient enhanced by the metamaterial structure was 424 cm/GW, which is 1.2 × 103 times higher than that of Si. Using this metamaterial waveguide-based nonlinear optical activation function, we achieved a high inference accuracy of 98.36% in the handwritten character recognition task, comparable to that obtained with the ReLU function as the activation function. Therefore, our approach can contribute to the realization of more power-efficient and compact all-optical neural networks.

Tensor networks for black hole interiors: non-isometries, quantum extremal surfaces, and wormholes

We use hyperbolic tensor networks to construct a holographic map for black hole interiors that adds a notion of locality to the non-isometric codes proposed by Akers, Engelhardt, Harlow, Penington, and Vardhan. We use tools provided by these networks to study the relationship between non-isometries and quantum extremal surfaces behind the horizon. Furthermore, we introduce a limited notion of dynamics for these interior tensor networks based on the qudit models introduced by Akers et al., and study the evolution of quantum extremal surfaces in an evaporating black hole. We also find a tensor network description of a wormhole connecting the black hole interior to the radiation, providing a mechanism for interior states and operators to be encoded in the radiation after the Page time. As a particular case, we construct a tensor network realization of the backwards-forwards maps recently proposed to incorporate non-trivial effective dynamics in dynamical constructions of these non-isometric black hole codes.

Coupled EMAG-CFD-Thermal analysis of a novel SST system utilizing forced-liquid cooling with biodegradable dielectric fluid

Solid state transformers (SSTs) are at the forefront of emerging technologies, enhancing distribution and transmission networks. These transformers facilitate the development of highly controllable and robust smart grids, seamlessly integrating renewable energy sources and accommodating connections for both alternating current and direct current components/lines at medium or low voltage levels. This research paper introduces an innovative SST incorporating the inductive power transfer (IPT) technology, currently under development within SSTAR, a Horizon Europe project. This novel device operates at a frequency of 50 kHz and provides with an output power of 50 kW. The IPT system is submerged in an ester-based, biodegradable dielectric fluid, providing effective cooling and insulation characteristics. The primary focus of this research is to highlight the guidelines for the optimal design of the forced liquid cooling system of the SST module. The analysis employs a multi-faceted simulation strategy to calculate the cooling aspects of the transformer, utilizing the commercial software ANSYS. This involves 3D coupled Electromagnetic (EMAG) - Computational Fluid Dynamics (CFD) simulations, as well as stand-alone CFD and thermal FEA simulations, ensuring robust cross-verification of the obtained results. This novel SST exhibits improved thermal performance, achieving a hot-spot temperature of 58 °C, the lowest value found in similar research studies. The findings underscore the efficiency and robustness of the proposed design, marking a significant step towards the realization of active and environmentally-friendly electrical networks.

Solute transport characteristics of columnar volumetric contraction networks with mega column structure and aperture variability

Numerical simulations explore for the first time the role of mega columns and aperture variability on particle transport through mature volumetric contraction networks as informed by a unique synthesis of network propagation and maturity. Columnar fracture patterns are generated by updating a series of Voronoi centers to the midpoint of a generated polygon over many iterations, creating 250 network realizations. A DFN simulator solves for fluid flow and tracks conservative particle trajectories within each network. Dominant fracture attributes impacting fluid flow and solute transport in volumetric contraction networks are fracture orientation, density, and aperture/transmissivity. Ensemble plume snapshots generated by networks with equal fracture transmissivity define a baseline-level of dispersion that is solely attributed to network structure and connectivity. Longitudinal and transverse dispersion increase and the center of plume mass becomes delayed relative to the baseline case when fracture transmissivity is varied according to a lognormal distribution. The incorporation of highly-transmissive, large-aperture mega column fractures leads to plume snapshots with a more pronounced leading edge and an order of magnitude faster average breakthrough times. The breakthrough curves contain three peaks reflecting contrasting transport pathways in which particles are: (i) initially placed in mega column fractures and remain in these features until exiting the model domain, (ii) initially placed into small column fractures, incur additional time to migrate and enter a mega column fracture, and remain within those mega columns, and (iii) initially placed in small column fractures and remain in these fractures. Incorporating variability in fracture transmissivity for both small column and mega column fractures disrupts the binary distinction between small column and mega column fracture velocities and leads to dispersed breakthroughs over long time scales with a single peak. These results demonstrate that preferential flow paths emerge in volumetric contraction networks due to contracts in fracture transmissivity, not fracture connectivity as observed in tectonic networks.

Design and Performance Analysis of Compact Printed Ridge Gap Waveguide Phase Shifters for Millimeter-Wave Systems.

This paper introduces compact Printed Ridge Gap Waveguide (PRGW) phase shifters tailored for millimeter-wave applications, with a focus on achieving wide operating bandwidth, and improved matching and phase balance compared to single-layer technology. This study proposes a unique approach to achieve the required phase shift in PRGW technology, which has not been previously explored. This study also introduces a novel analytical approach to calculate the cutoff frequency and propagation constant of the PRGW structure, a method not previously addressed. Furthermore, the utilization of multi-layer PRGW technology enables the realization of multi-layer beamforming networks without crossing, thereby supporting wideband operation in a compact size. The proposed design procedure enables the realization of various phase shift values ranging from 0∘ to 135∘ over a broad frequency bandwidth centered at 30 GHz. A 45-degree phase shifter is fabricated and tested, demonstrating a 10 GHz bandwidth (approximately 33% fractional bandwidth) from 25 GHz to 35 GHz. Throughout the operating bandwidth, the phase balance remains within 45 ± 5∘, with a deep matching level of -20 dB. The proposed phase shifter exhibits desirable characteristics, such as compactness, low loss, and low dispersion, making it a suitable choice for millimeter-wave applications, including beyond 5G (B5G) and 6G wireless communications.

Optimized CNN and adaptive RBFNN for channel estimation and hybrid precoding approaches for multi user millimeter wave massive MIMO

ABSTRACT The millimetre-wave (mmWave) communication satisfies the demand for high data rates due to the characteristic of wide bandwidth. Using massive multiple-input multiple-output (MIMO) technology, a significant propagation loss of mmWave communication is effectively compensated. However, it is challenging to provide a specialised radio frequency chain for each antenna due to the constrained physical area with closely spaced antennas and prohibitive power consumption in mmWave massive MIMO systems. This paper presents novel approaches for effective channel estimation and hybrid precoding in mmWave communication systems. To address the challenges of channel estimation, a convolutional neural network (CNN) is utilised, and network parameters are optimised using enhanced whale optimization algorithm (EWOA). The proposed CNN-based channel estimation method aims to accurately estimate the channel in mmWave systems with enhanced efficiency and reduced complexity. By training CNN using EWOA optimisation algorithm, the network parameters are fine-tuned to improve accuracy and generalisation capability of channel estimation process. Furthermore, hybrid precoding is achieved using adaptive radial-basis function neural networks (adaptive RBFNNs) which enables efficient precoding while minimising complexity. Moreover, the adaptive RBFNN approach determines the optimal precoding weights based on channel state information, resulting in a improved performance and a reduced computational overhead. The performance analysis is validated using the MATLAB/Simulink software and offers to provide effectual and reliable mmWave communication systems, facilitating the realisation of high-speed and high-capacity wireless networks.

Micrometric thermal electronic nose able to detect and quantify individual gases in a mixture

Recent urbanization and environmental problems urge for networks of sensors that can monitor air quality. Small, inexpensive, and smart sensors are one of the key components enabling the realization of such networks. Chemoresistive sensors are the ideal candidate, but they greatly lack selectivity, and for this reason, they are usually combined in arrays to create electronic noses, whose dimensions, however, make them not miniaturizable and cannot be integrated into portable devices. To overcome this inconvenience, we present a thermal electronic nose consisting of identical resistive sensors working at different temperatures so that the whole device is simple to make and tiny. The device contains two sensor arrays based on tin oxide nanowires decorated with Ag and Pt nanoparticles, respectively. The five sensors in each array are identical, but their response is differentiated by different temperatures locally generated by an on-chip integrated heater. This innovative approach allows the tiny array of five sensors together with the integrated heater to occupy only approximately 50 × 200 μm2 and consume only 120 μW. The tiny and portable device can estimate the concentration of H2 and NH3 in a mixture with a root mean square error of 6.1 ppm and 13.3 ppm, respectively, and it still works well after two months. The performance analysis of the double partial least squares regression used for concentration estimation also allows for feedback on which sensors are the most sensitive to which gas so that the electronic nose can be engineered for specific applications using the most suitable sensors. The size of the thermal electronic nose allows it to be integrated into portable and wearable devices, and its performance makes it suitable for any gas detection application. For example, a smartphone with an integrated sensor could carry out breath analysis and act as medical pre-screening or be used to evaluate the freshness of agri-food products in a rapid and non-invasive way.

Low-frequency Raman active modes of twisted bilayer MoS2

We study the low-frequency Raman active modes of twisted bilayer MoS2 for several twist angles using a force-field approach and a parametrized bond polarizability model. We show that twist angles near high-symmetry stacking configurations exhibit stacking frustration that leads to significant buckling of the moiré superlattice. We find that atomic relaxation due to the twist is of prime importance. The periodic displacement of the Mo atoms shows the realization of a soliton network, and in turn, leads to the emergence of a number of frequency modes not seen in the high-symmetry stacking systems. Some of the modes are only seen in the XZ Raman polarization setup while others are seen in the XY setup. The symmetry of the normal modes, and how this affects the Raman tensors is examined in detail.

Multi-heterointerface integration in nitrogen-doped Ti3C2Tx@Fe3O4 nanocomposites for superior microwave absorption

Engineering a lightweight and thin microwave absorber through the design of multiple heterogeneous interfaces represents a significant challenge. Here, we synthesized a nitrogen (N)-doped Ti3C2Tx@Fe3O4 (N–Ti3C2Tx@Fe3O4) nanocomposite with a sandwich-like structure through a solvothermal approach using hydrazine hydrate as the N donor. The numerous Ti3C2Tx electron transfer pathways of Ti3C2Tx enable the realization of a hierarchical conduction network within N–Ti3C2Tx@Fe3O4. The introduction of defects through N doping and the integration of Fe3O4 with Ti3C2Tx are pivotal for enhancing the dipole polarization. The heterogeneous interfaces in N–Ti3C2Tx@Fe3O4, coupled with Fe3O4 components, substantially boost the electromagnetic wave (EMW) absorption performance of the composite. An optimal N integration and the loading of Fe3O4 in Ti3C2Tx result in an improved impedance matching of N–Ti3C2Tx@Fe3O4. The synergistic effect of the improved impedance matching and the strong EMW attenuation capability endow N–Ti3C2Tx@Fe3O4 with enhanced microwave absorption (MA) properties. Specifically, at a thickness of 1.04 mm, the composite exhibits the effective absorption bandwidth (3.92 GHz; 14.08–18.00 GHz), surpassing that of Ti3C2Tx. At a thickness of 0.98 mm, it exhibits a maximum reflection loss of −40.28 dB at 17.25 GHz, which substantially exceeds that of Ti3C2Tx. Furthermore, it shows a maximum Radar Cross Section reduction of 21.32 dB cm2. A cotton fabric was coated with N–Ti3C2Tx@Fe3O4, and it was found to exhibit enhanced heat conduction and dissipation, at rates 3.83 and 4.64 times greater than those of uncoated cotton, respectively. These results highlight the considerable potential of N–Ti3C2Tx@Fe3O4 nanocomposites for use in far-field applications. It is envisaged that the exceptional MA and thermal management properties of N–Ti3C2Tx@Fe3O4 nanocomposites will provide a novel avenue for crafting lightweight and thin EMW absorbing materials.

Hybrid synaptic structure for spiking neural network realization

Neural networks and neuromorphic computing represent fundamental paradigms as alternative approaches to Von-Neumann-based implementations, advancing in the applications of deep learning and machine vision. Nonetheless, conventional semiconductor circuits encounter challenges in achieving ultra-fast processing speed and low power consumption due to their dissipative properties. Conversely, single flux quantum circuits exhibit inherent spiking behavior, showcasing their characteristics as a promising candidate for spiking neural networks (SNNs). In this work, we present a compact hybrid synapse circuit to mimic the biological interconnect functionality, enabling the weighting operations for excitatory and inhibitory impulses. Additionally, the proposed structure facilitates input accumulation, which is performed before the activation function. In the experiments, our synaptic structure interfaces with a soma circuit fabricated using a commercial Nb process, underscoring its compatibility and supporting its potential for integration into efficient neural network architectures. The weight value on the synapse is configurable by utilizing cryo-CMOS circuits, providing adaptability to the inference networks. We’ve successfully designed, fabricated, and partially tested the JJ-Synapse within our cryocooler system, enabling high-speed inference implementation for SNNs.

Multiplexable all-optical nonlinear activator for optical computing

As an alternative solution to surpass electronic neural networks, optical neural networks (ONNs) offer significant advantages in terms of energy consumption and computing speed. Despite the optical hardware platform could provide an efficient approach to realizing neural network algorithms than traditional hardware, the lack of optical nonlinearity limits the development of ONNs. Here, we proposed and experimentally demonstrated an all-optical nonlinear activator based on the stimulated Brillouin scattering (SBS). Utilizing the exceptional carrier dynamics of SBS, our activator supports two types of nonlinear functions, saturable absorption and rectified linear unit (Relu) models. Moreover, the proposed activator exhibits large dynamic response bandwidth (∼11.24 GHz), low nonlinear threshold (∼2.29 mW), high stability, and wavelength division multiplexing identities. These features have potential advantages for the physical realization of optical nonlinearities. As a proof of concept, we verify the performance of the proposed activator as an ONN nonlinear mapping unit via numerical simulations. Simulation shows that our approach achieves comparable performance to the activation functions commonly used in computers. The proposed approach provides support for the realization of all-optical neural networks.

Modeling the performance and bandwidth of single-atom adiabatic quantum memories

Quantum memories are essential for quantum repeaters, which will form the backbone of the future quantum internet. Such memory can capture a signal state for a controllable amount of time, after which this state can be retrieved. In this work, we theoretically investigated how atomic material and engineering parameters affect the performance and bandwidth of a quantum memory. We have applied a theoretical model for quantum memory operation based on the Lindblad master equation and adiabatic quantum state manipulation. The materials’ properties and their uncertainty are evaluated to determine the performance of Raman-type quantum memories by showcasing two defects in two-dimensional hexagonal boron nitride. We have derived a scheme to calculate the signal bandwidth based on the material parameters as well as the maximum efficiency that can be realized. The bandwidth depends on four factors: the signal photon frequency, the dipole transition moments in the electronic structure, the cavity volume, and the strength of the external control electric field. As our scheme is general and independent of materials, it can be applied to many other quantum materials with a suitable three-level structure. We, therefore, provided a promising route for designing and selecting materials for quantum memories. Our work is, therefore, an important step toward the realization of a large-scale quantum network.