Tunable Network Research Articles

6G autonomous radio access network empowered by artificial intelligence and network digital twin

AbstractThe sixth-generation (6G) mobile network implements the social vision of digital twins and ubiquitous intelligence. Contrary to the fifth-generation (5G) mobile network that focuses only on communications, 6G mobile networks must natively support new capabilities such as sensing, computing, artificial intelligence (AI), big data, and security while facilitating Everything as a Service. Although 5G mobile network deployment has demonstrated that network automation and intelligence can simplify network operation and maintenance (O&M), the addition of external functionalities has resulted in low service efficiency and high operational costs. In this study, a technology framework for a 6G autonomous radio access network (RAN) is proposed to achieve a high-level network autonomy that embraces the design of native cloud, native AI, and network digital twin (NDT). First, a service-based architecture is proposed to re-architect the protocol stack of RAN, which flexibly orchestrates the services and functions on demand as well as customizes them into cloud-native services. Second, a native AI framework is structured to provide AI support for the diverse use cases of network O&M by orchestrating communications, AI models, data, and computing power demanded by AI use cases. Third, a digital twin network is developed as a virtual environment for the training, pre-validation, and tuning of AI algorithms and neural networks, avoiding possible unexpected losses of the network O&M caused by AI applications. The combination of native AI and NDT can facilitate network autonomy by building closed-loop management and optimization for RAN.

Computational Insights into Tunable Reversible Network Materials: Accelerated ReaxFF Kinetics of Furan-Maleimide Diels-Alder Reactions for Self-Healing and Recyclability.

In this study, ReaxFF molecular dynamics simulations were benchmarked and used to study the relative kinetics of the retro Diels-Alder reaction between furan and N-methylmaleimide. This reaction is very important for the creation of polymer networks with self-healing and recyclable properties, since they can be used as reversible linkers in the network. So far, the reversible Diels-Alder reaction has not yet been studied by using reactive molecular dynamics simulations. This work is, thus, the first step in simulating a covalent adaptable network (CAN) using Diels-Alder reactions as reversible linkers. For both endo and exo, the bond breaking in 40 product molecules was simulated using the bond boost method and the endo/exo ratio was evaluated. This ratio was benchmarked against density functional theory (DFT) and experimental results for a changing set of bond boost parameters. Given their importance to understand how the CAN performs, the effect of the addition of a polymer backbone and the effect of temperature were successfully simulated using our newly parametrized reactive force field.

Radiofrequency Coil Tuning Using Matching Circuit Co-Simulation with Magnetic Resonance Integral Equations

Abstract The proposed extension to the conventional circuit co-simulation (CCS) method reliably and efficiently introduces arbitrary matching networks into CCS routines and is here dubbed as matching circuit co-simulation (MCCS). It further achieves coil tuning for resonance and an optimum scattering parameters condition that resembles a fully matched coil. Combined with magnetic resonance integral equations (MARIE), the proposed MCCS further accurately produces a full EM simulation of the coils. We first validated MARIE with the MCCS routine to use it to model simple single-channel coils and parallel transmit (pTx) head arrays coupled with Duke, a digital human body model. Having successfully validated the MCCS routine, we swept through both tuning and matching network parameters to optimize the coil. While optimizing for all parameters simultaneously is a viably efficient option for coil models of low complexity, these models are not realistic, and we found that it becomes a highly computationally expensive problem to solve for more complex coil geometries, such as birdcage coils and pTx coils with higher number of channels, especially when shielding is included. Consequently, we separated tuning and matching into two steps, by first tuning the coil to the right frequency, and then introducing matching networks to enhance the power transmission into the body. We observed lower on-diagonal S-parameters after matching in all pTx coils, but 1 coil still had strong cross-coupling. Finally, electromagnetic fields obtained via MARIE for separate port excitation were successfully combined using MCCS.

Methodology for hyperparameter tuning of deep neural networks for efficient and accurate molecular property prediction

This paper presents a methodology of hyperparameter optimization (HPO) of deep neural networks for molecular property prediction (MPP). Most prior applications of deep learning to MPP have paid only limited attention to HPO, thus resulting in suboptimal values of predicted properties. To improve the efficiency and accuracy of deep learning models for MPP, we must optimize as many hyperparameters as possible and choose a software platform to enable the parallel execution of HPO. We compare the random search, Bayesian optimization, and hyperband algorithms, together with the Bayesian-hyperband combination within the software packages of KerasTuner and Optuna for HPO. We conclude that the hyperband algorithm, which has not been used in previous MPP studies, is most computationally efficient; it gives MPP results that are optimal or nearly optimal in terms of prediction accuracy. Based on our case studies, we recommend the use of the Python library KerasTuner for HPO.

Adaptive Static Equivalences for Active Distribution Networks With Massive Renewable Energy Integration: A Distributed Deep Reinforcement Learning Approach

Active distribution networks (ADNs) with 100% renewable energy sources (RESs) penetration are essential in the net-zeros emission power sector. As an efficient analysis tool, the equivalent model for ADNs (EMADNs) can be leveraged to expedite the lengthy analysis and preserve data privacy. Nevertheless, volatile RESs, along with their active management techniques, e.g., voltage regulation schemes (VRSs), deteriorate the fidelity of the developed EMADNs derived from historical data. This paper proposes an adaptive EMADN with tunable network scales and adaptive parameters to address the above challenges. A leaves-trimming network topological reduction algorithm is utilized to preserve the radial topology, the nodes of interest, and VRSs. Upon the derived network topology and components, a distributed Proximal Policy Optimization (DPPO)-based deep reinforcement learning agent is employed to adjust the parameters periodically to maintain the fidelity of EMADNs over time. The distributed training scheme extends the ordinary PPO to boost exploration and training efficiency by exploiting the multi-processors in a synchronous way. Case studies on the IEEE-33 bus and IEEE 123-bus ADNs with massive photovoltaic and wind generation demonstrate the superior accuracy and efficiency of the proposed agent-based EMADNs in various scenarios, especially when the production of RESs exceeds the load amount.

Deep learning coupled Bayesian inference method for measuring the elastoplastic properties of SS400 steel welds by nanoindentation experiment

Nanoindentation experiment has shown broad application prospects due to its ability to measure the mechanical properties of various materials at multiple scales. In this paper, a deep learning coupled Bayesian inverse approach is proposed for measuring the elastoplastic parameters of SS400 steel welds by nanoindentation experiment. The nanoindentation experiments were performed on the SS400 steel welds, including base metal (BM), weld zone (WZ), and heat affected zone (HAZ), and the experiment load–displacement (P-h) curves were obtained. The hyper-parameters tunable artificial neural network (ANN) was established to correlate elastoplastic parameters with indentation P-h curves. Based on Bayesian inference theory, the posterior density function for estimating the unknown material parameters was established. Transitional Markov chain Monte Carlo was used for sampling from the posterior density function, and the elastoplastic properties in different regions of SS400 steel welds were identified. The advantage of the established measuring method is that the hyper-parameters optimized ANN model can provide the very accurate forward relationship between material properties and indentation P-h curves. Besides, the inverse Bayesian framework can quantify the potential uncertainty of the identified elastoplastic parameters. The measured elastoplastic properties of the base metal of SS400 steel show good agreement with tensile experiment data, of which the maximum measuring error is less than 12%. The measured elastoplastic properties in WZ and HAZ are also proved to be effective. The uncertainty of the identified elastoplastic parameters of SS400 steel welds can be quantified by posterior marginal distribution, using Mean and Variance values. The results proved that the proposed inverse measuring method is reliable and effective.

Thermally Assisted Microfluidics to Produce Chemically Equivalent Microgels with Tunable Network Morphologies.

Although micron-sized microgels have become important building blocks in regenerative materials, offering decisive interactions with living matter, their chemical composition mostly significantly varies when their network morphology is tuned. Since cell behavior is simultaneously affected by the physical, chemical, and structural properties of the gel network, microgels with variable morphology but chemical equivalence are of interest. This work describes a new method to produce thermoresponsive microgels with defined mechanical properties, surface morphologies, and volume phase transition temperatures. A wide variety of microgels is synthesized by crosslinking monomers or star polymers at different temperatures using thermally assisted microfluidics. The diversification of microgels with different network structures and morphologies but of chemical equivalence offers a new platform of microgel building blocks with the ability to undergo phase transition at physiological temperatures. The method holds high potential to create soft and dynamic materials while maintaining the chemical composition for a wide variety of applications in biomedicine.

Facile synthesis and biomimetic amine-functionalization of chitosan foam for CO2 capture

A high-performance biomass-based adsorption materials could be the promising trend for CO2 capture and storage technology. However, the direct application of biomass-based porous materials as a CO2 adsorbent with enhanced performance is an emerging issue. Herein, a facile synthesis and a biomimetic strategy were combined to prepare amine-functionalized chitosan foam for CO2 capture, and then a porous biomass is achieved for the application on the environment protection field. Firstly, the chitosan foam was synthesized by the emulsion-templating method at room temperature. Depended on stabilizing n-octane in the chitosan hydrogel with Span 80, a tunable three-dimensional network porous structure was obtained. Subsequently, co-deposition with dopamine (DA) and polyethyleneimine (PEI) was applied to load abundant amine content on the surface of chitosan foam and thereby improving CO2 adsorption capacity. Finally, the as-prepared amine-functionalized chitosan foam exhibited the impressive adsorption capacity of 3.59 mmol/g at 333 K and atmospheric pressure, and the better adsorption selectivity and stability. The results extend the preparation approach of biomass porous materials, and also its application in CO2 adsorption technology.

Gelatin-based hydrogels with tunable network structure and mechanical property for promoting osteogenic differentiation

Osteoarthritis (OA) is a joint disease involving all joint components, including cartilage, calcified cartilage, and subchondral bone. The repair of osteochondral defects remains a significant challenge in orthopedics. Development of new strategies is essential for effective osteochondral injury repair. In this study, gelatin (Gel), polyethylene glycol diglycidyl ether (PEGDGE), hydroxyethyl cellulose (HEC) and chitosan (CS) were used to prepare semi-IPNs and IPNs hydrogels. Mechanical properties of Gel based hydrogels significantly improved with the semi-IPN and IPN structures. Tensile strength ranges from 238.7 KPa to 479.5 KPa, and its compressive strength ranges from 35.6 KPa to 112.7 KPa. Additionally, the stress relaxation rate increased with higher CS concentrations, ranging from 25 % to 35 %. The network structure of Gel-based hydrogels was a key factor in regulating stress relaxation. Viscoelasticity was adjusted by its network structures. Swelling and degradation behaviors of Gel based hydrogels were systematically investigated. Gel based hydrogels had good cytocompatibility. Both semi-IPN and IPN structures Gel based hydrogels could promote cell spreading and osteogenic differentiation. G10HEC1 and G10CS1 hydrogels show promise as candidates for osteochondral tissue regeneration, offering a new strategy for osteochondral tissue engineering.

Design of Doherty power amplifier with power back-off extension based on harmonic tuning and output combining network optimization

This paper proposes a Doherty power amplifier (DPA) design method based on harmonic tuning and output combining network (OCN) optimization to extend the power back-off (PBO) range. Firstly, based on the optimal harmonic load impedance, a harmonic tuning optimization method is used to design the harmonic tuning network of the amplifiers, effectively improving the efficiency at saturation and PBO. Secondly, the fundamental output matching network and the post-matching network are considered as an OCN. The S-parameters of the OCN are calculated and utilized in the DPA design. To further expand the PBO range, the carrier load impedance at PBO is determined by considering the relationship between PBO range and reflection coefficient. Finally, a multi-objective evolutionary algorithm combined with the theory of solution set is used to optimize the OCN design, simplifying the design process of the output matching network. For verification, a high-efficiency DPA operating at 1.68 GHz with a large PBO range of 9.5 dB is designed and measured. Results indicate that the proposed DPA achieves a saturated output power greater than 44 dBm, with a peak efficiency of up to 75.2%. The efficiencies are 68.5% and 61.0% at 6 and 9.5 dB back-off powers, respectively.

A hybrid deep learning skin cancer prediction framework

Skin cancer, a critical health concern, necessitates accurate early detection and classification to mitigate its impact. However, the limited availability of medical datasets and the challenge of optimizing learnable parameters in deep learning strategies for medical images pose significant hurdles. To address these challenges, we propose an innovative solution—a hybrid artificial intelligence (AI) framework for skin cancer prediction. This framework consists of two pivotal steps: firstly, a comprehensive skin cancer dataset is curated by combining diverse public datasets encompassing multiple disease types. This facilitates robust weight tuning of Convolutional Neural Networks (CNNs) by enhancing the diversity of the learning process. Subsequently, the optimization of deep learning is meticulously executed through the artificial Bee Colony (ABC) strategy, effectively mitigating the potential adverse effects of initial parameter randomness on AI model performance. Rigorous evaluation of the hybrid framework against eight prominent CNN models, including DenseNet121, InceptionResNetV2, and others, underscores its superior predictive capabilities. The optimized hybrid framework achieves exceptional results, boasting an accuracy of 93.04%, recall of 92.0%, precision of 93.0%, and an impressive F1-score of 93.12%. In summary, our proposed hybrid solution demonstrates a substantial advancement in multi-disease skin cancer classification. It outperforms existing AI models and holds promise as a pragmatic smart solution with far-reaching implications.

Electrofluids with Tailored Rheoelectrical Properties: Liquid Composites with Tunable Network Structures as Stretchable Conductors.

Flexible and stretchable electronics require both sensing elements and stretching-insensitive electrical connections. Conductive polymer composites and liquid metals are highly deformable but change their conductivity upon elongation and/or contain rare metals. Solid conductive composites are limited in mechanoelectrical properties and are often combined with macroscopic Kirigami structures, but their use is limited by geometrical restraints. Here, we introduce "Electrofluids", concentrated conductive particle suspensions with transient particle contacts that flow under shear that bridge the gap between classic solid composites and liquid metals. We show how Carbon Black (CB) forms large agglomerates when using incompatible solvents that reduce the electrical percolation threshold by 1 order of magnitude compared to more compatible solvents, where CB is well-dispersed. We analyze the correlation between stiffness and electrical conductivity to create a figure of merit of first electrofluids. Sealed elastomeric tubes containing different types of electrofluids were characterized under uniaxial tensile strain, and their electrical resistance was monitored. We found a dependency of the piezoresistivity with the solvent compatibility. Electrofluids enable the rational design of sustainable soft electronics components by simple solvent choice and can be used both as sensor and electrode materials, as we demonstrate.

Pore-networked membrane using linked metal-organic polyhedra for trace-level pollutant removal and detection in environmental water

The wide presence of pharmaceuticals and personal care products (PPCPs) in water is a major pollution concern even at the part per billion level, which urges their detection and removal. Current research emphasizes the use of microporous materials as adsorbents for pollutant removal but demonstrates the performance at a higher concentration than realistic environmental water due to the absence of efficient detection methods that can be coupled with the removal process. Here we report a pore-networked membrane (PNM) that can simultaneously remove and detect targeted trace-level PPCPs. These PNMs are designed by interconnecting adsorbents within polymer matrices, forming continuous, tunable porous networks that are accessible for PPCPs, thereby offering high selectivity and adsorption capacity. Evaluations across water samples containing 13 pollutants demonstrate the capability of PNMs to selectively adsorb PPCPs for removal and subsequently release them into analysis solution for detection, positioning their promising use in real water treatment workflow.

An Integrated Polysaccharide Hydrogel with Versatile Fluorescence Responses through Noncovalent Reformation of Gel Aggregation and for Bioimaging.

Functionalized hydrogels, with their unique and adaptable structures, have attracted significant attention in materials and biomaterials research. Fluorescent hydrogels are particularly noteworthy for their sensing capabilities and ability to mimic cellular matrices, facilitating cell infiltration and tracking of drug delivery. Structural elucidation of hydrogels is crucial for understanding their responses to stimuli such as the pH, temperature, and solvents. This study developed a fluorescent hydrogel by functionalizing chitosan with p-cresol-based quinazolinone aldehyde. Confocal microscopy revealed the hydrogel's intriguing fluorogenic properties. The hydrogel exhibited enhanced fluorescence and a tunable network morphology, influenced by the THF-water ratio. The study investigated the control of gel network reformation in different media and analyzed the fluorescence responses and structural changes of the sugar backbone and fluorophore. Proper selection of mixed solvents is essential for optimizing the hydrogel as a fluorescence probe for bioimaging. This hydrogel demonstrated greater swelling properties, making it highly suitable for drug delivery applications.

3.3-4.3 GHz efficient continuous class-F gallium nitride power amplifier based on simplified real frequency technique and harmonic tuning.

In order to implement the fifth generation (5G) communication system for a large number of users, the governments of many countries nominated the low 5G frequency band between 3.3 and 4.3 GHz. This paper proposes a wideband RFPA by designing the input matching network (MN) and output MN of the device using the simplified real frequency technique (SRFT) and the harmonic tuning network. The load-pull and source-pull is applied at multiple points for 100 MHz intervals over the bandwidth to obtain the optimum impedances at the output and input of the 10W Gallium Nitride (GaN) Cree CGH40010F device. To verify the design, the RFPA is simulated, and the performance is measured between 3.3 and 4.3 GHz. According to experimental findings, the measured drain efficiency (DE) throughout the whole bandwidth ranged from 57.5 to 67.5% at the output power of 40 dBm. Moreover, at the 1 dB compression point between 39.2 and 42.2 dBm output power, the drain efficiency (DE) achieves a high value of 81.2% with an output power of 42.2 dBm at a frequency of 3.3 GHz. The RFPA can obtain a maximum gain of 12.4 dB at 3.5 GHz. The linearity of the RFPA with a two-tone signal is measured and the value is less than -22 dBc all over the band.

Tunable Waterborne Aspartic Acid-Based Covalent Adaptable Networks with Internal Catalysis and Incorporation in Bioinspired Nanocomposites

Tunable Waterborne Aspartic Acid-Based Covalent Adaptable Networks with Internal Catalysis and Incorporation in Bioinspired Nanocomposites

A mechanically Robust, Damping, and High‐Temperature Tolerant Ion‐Conductive Elastomer for Noise‐Free Flexible Electronics

AbstractIon‐conductive elastomers capable of damping can significantly mitigate the interference caused by mechanical noise during data acquisition in wearable and biomedical devices. However, currently available damping elastomers often lack robust mechanical properties and have a narrow temperature range for effective damping. Here, precise modulation of weak to strong ion‐dipole interactions plays a crucial role in bolstering network stability and tuning the relaxation behavior of supramolecular ion‐conductive elastomers (SICEs). The SICEs exhibit impressive mechanical properties, including a modulus of 13.2 MPa, a toughness of 65.6 MJ m−3, and a fracture energy of 74.9 kJ m−2. Additionally, they demonstrate remarkable damping capabilities, with a damping capacity of 91.2% and a peak tan δ of 1.11. Furthermore, the entropy‐driven rearrangement of ion‐dipole interactions ensures the damping properties of the SICE remain stable even at elevated temperatures (18–200 °C, with tan δ > 0.3), making it the most thermally resistant damping elastomer reported to date. Moreover, the SICE proves effective in filtering out various noises during physiological signal detection and strain sensing, highlighting its vast potential in flexible electronics.

Unsupervised shape-and-texture-based generative adversarial tuning of pre-trained networks for carotid segmentation from 3D ultrasound images.

Vessel-wall volume and localized three-dimensional ultrasound (3DUS) metrics are sensitive to the change of carotid atherosclerosis in response to medical/dietary interventions. Manual segmentation of the media-adventitia boundary (MAB) and lumen-intima boundary (LIB) required to obtain these metrics is time-consuming and prone to observer variability. Although supervised deep-learning segmentation models have been proposed, training of these models requires a sizeable manually segmented training set, making larger clinical studiesprohibitive. We aim to develop a method to optimize pre-trained segmentation models without requiring manual segmentation to supervise the fine-tuningprocess. We developed an adversarial framework called the unsupervised shape-and-texture generative adversarial network (USTGAN) to fine-tune a convolutional neural network (CNN) pre-trained on a source dataset for accurate segmentation of a target dataset. The network integrates a novel texture-based discriminator with a shape-based discriminator, which together provide feedback for the CNN to segment the target images in a similar way as the source images. The texture-based discriminator increases the accuracy of the CNN in locating the artery, thereby lowering the number of failed segmentations. Failed segmentation was further reduced by a self-checking mechanism to flag longitudinal discontinuity of the artery and by self-correction strategies involving surface interpolation followed by a case-specific tuning of the CNN. The U-Net was pre-trained by the source dataset involving 224 3DUS volumes with 136, 44, and 44 volumes in the training, validation and testing sets. The training of USTGAN involved the same training group of 136 volumes in the source dataset and 533 volumes in the target dataset. No segmented boundaries for the target cohort were available for training USTGAN. The validation and testing of USTGAN involved 118 and 104 volumes from the target cohort, respectively. The segmentation accuracy was quantified by Dice Similarity Coefficient (DSC), and incorrect localization rate (ILR). Tukey's Honestly Significant Difference multiple comparison test was employed to quantify the difference of DSCs between models and settings, where was considered statisticallysignificant. USTGAN attained a DSC of % in LIB and % in MAB, improving from the baseline performance of % in LIB (p ) and % in MAB (p ). Our approach outperformed six state-of-the-art domain-adaptation models (MAB: , LIB: ). The proposed USTGAN also had the lowest ILR among the methods compared (LIB: 2.5%, MAB: 1.7%). Our framework improves segmentation generalizability, thereby facilitating efficient carotid disease monitoring in multicenter trials and in clinics with less expertise in 3DUSimaging.

Optimizing Vehicle Detection and Tracking Efficiency Through YOLO-Based Multi-Objective Approach

In traffic management, a careful detection of cars plays and monitors a vital role in ensuring safety and efficiency. This paper proposes a way to enhance the discovery of vehicles by dividing roads into remote areas and closing. The newly developed approach aims to define a vehicle and calculate vehicles in both fields. Yolov5M network is easy to discover and localize vehicles, exceed the traditional speed tracking methods and enable microorganisms. The proposed form achieves a higher accuracy of 0.833, outperform the Yolov5 Standard Form of 0.67. It also demonstrates the performance enhancement in the results of summons and mean average precision by keeping the training parameters reduction trend the same. Technically speaking, the split of the roads surface is divided into sections using the latest retail techniques and the fine tuning of the Yolov5M network to positively affect the detection and classification of vehicles.

Effect of Micellar Morphology on the Temperature-Induced Structural Evolution of ABC Polypeptoid Triblock Terpolymers into Two-Compartment Hydrogel Network.

We investigated the temperature-dependent structural evolution of thermoreversible triblock terpolypeptoid hydrogels, namely poly(N-allyl glycine)-b-poly(N-methyl glycine)-b-poly(N-decyl glycine) (AMD), using small-angle neutron scattering (SANS) with contrast matching in conjunction with X-ray scattering and cryogenic transmission electron microscopy (cryo-TEM) techniques. At room temperature, A100M101D10 triblock terpolypeptoids self-assemble into core-corona-type spherical micelles in aqueous solution. Upon heating above the critical gelation temperature (T gel), SANS analysis revealed the formation of a two-compartment hydrogel network comprising distinct micellar cores composed of dehydrated A blocks and hydrophobic D blocks. At T ≳ T gel, the temperature-dependent dehydration of A block further leads to the gradual rearrangement of both A and D domains, forming well-ordered micellar network at higher temperatures. For AMD polymers with either longer D block or shorter A block, such as A101M111D21 and A43M92D9, elongated nonspherical micelles with a crystalline D core were observed at T < T gel. Although these enlarged crystalline micelles still undergo a sharp sol-to-gel transition upon heating, the higher aggregation number of chains results in the immediate association of the micelles into ordered aggregates at the initial stage, followed by a disruption of the spatial ordering as the temperature further increases. On the other hand, fiber-like structures were also observed for AMD with longer A block, such as A153M127D10, due to the crystallization of A domains. This also influences the assembly pathway of the two-compartment network. Our findings emphasize the critical impact of initial micellar morphology on the structural evolution of AMD hydrogels during the sol-to-gel transition, providing valuable insights for the rational design of thermoresponsive hydrogels with tunable network structures at the nanometer scale.