Dynamic Architecture Research Articles

Time-resolved, deuterium-based fluxomics uncovers the hierarchy and dynamics of sugar processing by Pseudomonas putida

Pseudomonas putida, a microbial host widely adopted for metabolic engineering, processes glucose through convergent peripheral pathways that ultimately yield 6-phosphogluconate. The periplasmic gluconate shunt (PGS), composed by glucose and gluconate dehydrogenases, sequentially transforms glucose into gluconate and 2-ketogluconate. Although the secretion of these organic acids by P. putida has been extensively recognized, the mechanism and spatiotemporal regulation of the PGS remained elusive thus far. To address this challenge, we adopted a dynamic 13C- and 2H-metabolic flux analysis strategy, termed D-fluxomics. D-fluxomics demonstrated that the PGS underscores a highly dynamic metabolic architecture in glucose-dependent batch cultures of P. putida, characterized by hierarchical carbon uptake by the PGS throughout the cultivation. Additionally, we show that gluconate and 2-ketogluconate accumulation and consumption can be solely explained as a result of the interplay between growth rate-coupled and decoupled metabolic fluxes. As a consequence, the formation of these acids in the PGS is inversely correlated to the bacterial growth rate—unlike the widely studied overflow metabolism of Escherichia coli and yeast. Our findings, which underline survival strategies of soil bacteria thriving in their natural environments, open new avenues for engineering P. putida towards efficient, sugar-based bioprocesses.

A Prediction-Based Route Guidance Method Toward Intelligent and Green Transportation System

For the application of intelligent and green transportation systems (e.g., autonomous driving), traffic congestion is a severe challenge. So far, when traffic congestion is perceived for a route, a common solution is searching for another congestion-free route. However, it is observed that not all congestion should be tackled with re-routing since the extra overhead (e.g., travel time, fuel consumption, and CO2 emission) caused by specific congestion might be lower than that of re-routing. Against this backdrop, a Prediction-based Route Guidance Method (PRGM) is proposed for intelligent and green transportation systems. To begin with, PRGM involves a novel hybrid and dynamic system architecture based on the collaboration of vehicle clusters and the cloud platform. Notably, a backup mechanism between adjacent cluster heads is designed to avoid the problem that the data might be lost during dynamic clustering. Furthermore, PRGM involves a novel traffic congestion control strategy, which is based on four procedures: perception about traffic congestion with three indexes (i.e., speed index, dense index, acceleration index); judgment about congestion type with four defined congestion types; prediction about congestion duration considering the formation of congestion (i.e., why and how the congestion is formed); route planning about vehicles considering congestion duration and the extra time overhead of re-routing. Simulations are performed, and they show that the proposed PRGM not only can perceive traffic congestion more precisely and timely but also can reduce the travel time, fuel consumption, and CO2 emission of vehicles.

Optimizing Telematics Network Performance through Resource Virtualization in a Disruptive Environment: The Case of the IP/MPLS Core Network

We offer a security solution to considerably reduce latency in an IP network by virtualizing the IP/MPLS core network. It consists of adapting a virtualization method to a complex IP network, presenting the simulation of the implementation of this virtualization and the modifications to be made to certain aspects of the code of the solution. These modifications would take into account the key performance indicators of the network in order to guarantee its security and the transmission through very wide bands of data. To do this, we use Software Defined Network (SDN) technology. It allows us to have an emergent, scalable, dynamic, secure, laudable and adaptable network architecture, making it suitable for today's high bandwidth applications and IT services. This architecture decouples network control and digital data transfer functions, making network control directly programmable and the underlying infrastructure abstracted to network applications and services. After describing the soft failover to a virtualized network, we present the new architecture that describes the separation of the control and data planes of the core of the IP/MPLS network of the Autonomous Port of Kribi (PAK) in Cameroon, as part of our research work. We will then present the aspects in which modifying the code would contribute to improving one of the key qualities of service, namely latency in the heart of the network. We go from latencies above 100 ms to latencies below 1 ms; finally we recommend the approach for a continuous modification of the code with a view to optimizing the performance of the network in a continuous process for the reduction of its latency.

Adaptive timing in a dynamic field architecture for natural human–robot interactions

A close temporal coordination of actions and goals is crucial for natural and fluent human–robot interactions in collaborative tasks. How to endow an autonomous robot with a basic temporal cognition capacity is an open question. In this paper, we present a neurodynamics approach based on the theoretical framework of dynamic neural fields (DNF) which assumes that timing processes are closely integrated with other cognitive computations. The continuous evolution of neural population activity towards an attractor state provides an implicit sensation of the passage of time. Highly flexible sensorimotor timing can be achieved through manipulations of inputs or initial conditions that affect the speed with which the neural trajectory evolves. We test a DNF-based control architecture in an assembly paradigm in which an assistant hands over a series of pieces which the operator uses among others in the assembly process. By watching two experts, the robot first learns the serial order and relative timing of object transfers to subsequently substitute the assistant in the collaborative task. A dynamic adaptation rule exploiting a perceived temporal mismatch between the expected and the realized transfer timing allows the robot to quickly adapt its proactive motor timing to the pace of the operator even when an additional assembly step delays a handover. Moreover, the self-stabilizing properties of the population dynamics support the fast internal simulation of acquired task knowledge allowing the robot to anticipate serial order errors.

Dynamic Heterogeneous Graph Attention Neural Architecture Search

Dynamic heterogeneous graph neural networks (DHGNNs) have been shown to be effective in handling the ubiquitous dynamic heterogeneous graphs. However, the existing DHGNNs are hand-designed, requiring extensive human efforts and failing to adapt to diverse dynamic heterogeneous graph scenarios. In this paper, we propose to automate the design of DHGNN, which faces two major challenges: 1) how to design the search space to jointly consider the spatial-temporal dependencies and heterogeneous interactions in graphs; 2) how to design an efficient search algorithm in the potentially large and complex search space. To tackle these challenges, we propose a novel Dynamic Heterogeneous Graph Attention Search (DHGAS) method. Our proposed method can automatically discover the optimal DHGNN architecture and adapt to various dynamic heterogeneous graph scenarios without human guidance. In particular, we first propose a unified dynamic heterogeneous graph attention (DHGA) framework, which enables each node to jointly attend its heterogeneous and dynamic neighbors. Based on the framework, we design a localization space to determine where the attention should be applied and a parameterization space to determine how the attention should be parameterized. Lastly, we design a multi-stage differentiable search algorithm to efficiently explore the search space. Extensive experiments on real-world dynamic heterogeneous graph datasets demonstrate that our proposed method significantly outperforms state-of-the-art baselines for tasks including link prediction, node classification and node regression. To the best of our knowledge, DHGAS is the first dynamic heterogeneous graph neural architecture search method.

Leveraging Old Knowledge to Continually Learn New Classes in Medical Images

Class-incremental continual learning is a core step towards developing artificial intelligence systems that can continuously adapt to changes in the environment by learning new concepts without forgetting those previously learned. This is especially needed in the medical domain where continually learning from new incoming data is required to classify an expanded set of diseases. In this work, we focus on how old knowledge can be leveraged to learn new classes without catastrophic forgetting. We propose a framework that comprises of two main components: (1) a dynamic architecture with expanding representations to preserve previously learned features and accommodate new features; and (2) a training procedure alternating between two objectives to balance the learning of new features while maintaining the model’s performance on old classes. Experiment results on multiple medical datasets show that our solution is able to achieve superior performance over state-of-the-art baselines in terms of class accuracy and forgetting.

Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning

The dynamic expansion architecture is becoming popular in class incremental learning, mainly due to its advantages in alleviating catastrophic forgetting. However, task confu- sion is not well assessed within this framework, e.g., the discrepancy between classes of different tasks is not well learned (i.e., inter-task confusion, ITC), and certain prior- ity is still given to the latest class batch (i.e., old-new con- fusion, ONC). We empirically validate the side effects of the two types of confusion. Meanwhile, a novel solution called Task Correlated Incremental Learning (TCIL) is pro- posed to encourage discriminative and fair feature utilization across tasks. TCIL performs a multi-level knowledge distil- lation to propagate knowledge learned from old tasks to the new one. It establishes information flow paths at both fea- ture and logit levels, enabling the learning to be aware of old classes. Besides, attention mechanism and classifier re- scoring are applied to generate more fair classification scores. We conduct extensive experiments on CIFAR100 and Ima- geNet100 datasets. The results demonstrate that TCIL con- sistently achieves state-of-the-art accuracy. It mitigates both ITC and ONC, while showing advantages in battle with catas- trophic forgetting even no rehearsal memory is reserved. Source code: https://github.com/YellowPancake/TCIL.

Photoperiod and gravistimulation-associated Tiller Angle Control 1 modulates dynamic changes in rice plant architecture.

TAC1 is involved in photoperiodic and gravitropic responses to modulate rice dynamic plant architecture likely by affecting endogenous auxin distribution, which could explain TAC1 widespread distribution in indica rice. Plants experience a changing environment throughout their growth, which requires dynamic adjustments of plant architecture in response to these environmental cues. Our previous study demonstrated that Tiller Angle Control 1 (TAC1) modulates dynamic changes in plant architecture in rice; however, the underlying regulatory mechanisms remain largely unknown. In this study, we show that TAC1 regulates plant architecture in an expression dose-dependent manner, is highly expressed in stems, and exhibits dynamic expression in tiller bases during the growth period. Photoperiodic treatments revealed that TAC1 expression shows circadian rhythm and is more abundant during the dark period than during the light period and under short-day conditions than under long-day conditions. Therefore, it contributes to dynamic plant architecture under long-day conditions and loose plant architecture under short-day conditions. Gravity treatments showed that TAC1 is induced by gravistimulation and negatively regulates shoot gravitropism, likely by affecting auxin distribution. Notably, the tested indica rice containing TAC1 displayed dynamic plant architecture under natural long-day conditions, likely explaining the widespread distribution of TAC1 in indica rice. Our results provide new insights into TAC1-mediated regulatory mechanisms for dynamic changes in rice plant architecture.

Dynamic chromatin architecture of the porcine adipose tissues with weight gain and loss

Using an adult female miniature pig model with diet-induced weight gain/weight loss, we investigated the regulatory mechanisms of three-dimensional (3D) genome architecture in adipose tissues (ATs) associated with obesity. We generated 249 high-resolution in situ Hi-C chromatin contact maps of subcutaneous AT and three visceral ATs, analyzing transcriptomic and chromatin architectural changes under different nutritional treatments. We find that chromatin architecture remodeling underpins transcriptomic divergence in ATs, potentially linked to metabolic risks in obesity development. Analysis of chromatin architecture among subcutaneous ATs of different mammals suggests the presence of transcriptional regulatory divergence that could explain phenotypic, physiological, and functional differences in ATs. Regulatory element conservation analysis in pigs and humans reveals similarities in the regulatory circuitry of genes responsible for the obesity phenotype and identified non-conserved elements in species-specific gene sets that underpin AT specialization. This work provides a data-rich tool for discovering obesity-related regulatory elements in humans and pigs.

Dynamic Architecture as an Approach to Educational Building Development

Dynamic Architecture as an Approach to Educational Building Development

Detection of DDOS Attack using Decision Tree Classifier in SDN Environment

Software Defined Networking (SDN) is a dynamic architecture that employs a variety of applications for making networks more adaptable and centrally controlled. It is easy to attack the entire network in SDN because the control plane and data plane are separated. DDoS attack is major danger to SDN service providers because it can shut down the entire network and stop services to all customers at any time. One of the key flaws of most SDN architectures is lack of susceptibility to DDoS attacks with its types like TCP flooding, UDP flooding, SYN flooding, ICMP flooding and DHCP flooding for detecting those kinds of attacks. The machine learning algorithms are widely used in recent years to identify DDoS attacks. This research utilizes Decision Tree Classifier for detection and classification of DDoS attacks on SDN. The Forward Feature Selection technique is also used in the research to select the best features from the dataset and from that dataset the data are employed to train and test the model by Decision Tree Classifier Algorithm. The decision Tree Classifier technique is a supervised method used to forecast desired values of observations using rudimentary machine learning decision rules derived from training data. Based on the accuracy of decision tree techniques, in future, a hybrid learning model will be designed for detecting the Distributed Denial of Services in an SDN environment with high accuracy and a low false negative rate.

Neural Network Model-Based Control for Manipulator: An Autoencoder Perspective.

Recently, neural network model-based control has received wide interests in kinematics control of manipulators. To enhance learning ability of neural network models, the autoencoder method is used as a powerful tool to achieve deep learning and has gained success in recent years. However, the performance of existing autoencoder approaches for manipulator control may be still largely dependent on the quality of data, and for extreme cases with noisy data it may even fail. How to incorporate the model knowledge into the autoencoder controller design with an aim to increase the robustness and reliability remains a challenging problem. In this work, a sparse autoencoder controller for kinematic control of manipulators with weights obtained directly from the robot model rather than training data is proposed for the first time. By encoding and decoding the control target though a new dynamic recurrent neural network architecture, the control input can be solved through a new sparse optimization formulation. In this work, input saturation, which holds for almost all practical systems but usually is ignored for analysis simplicity, is also considered in the controller construction. Theoretical analysis and extensive simulations demonstrate that the proposed sparse autoencoder controller with input saturation can make the end-effector of the manipulator system track the desired path efficiently. Further performance comparison and evaluation against the additive noise and parameter uncertainty substantiate robustness of the proposed sparse autoencoder manipulator controller.

Towards an understanding of reliability of software-intensive systems-of-systems

Large-scale software-intensive Systems-of-Systems (SoS) have become present in several critical domains and have sometimes depended on diverse trending technologies, such as cloud computing and machine learning. At the same time, the SoS dynamic architecture makes it difficult to assure SoS reliability leading to diverse studies with specific solutions, while the need for a shared view of what precisely SoS reliability refers to still exists. The main contribution of this article is to go towards an understanding of SoS reliability. We present a conceptual model whose concepts as well as their definitions and relationships were defined by systematically examining the literature of the field. We surveyed 36 practitioners and researchers regarding ambiguity, explanatory power, parsimony, generality, and utility of our model. Next, we adjusted our model according to their contribution. We reach a conceptual model containing 29 concepts and their relationships that help to comprehend SoS reliability. In addition, we provided a glossary with a definition of each concept of our conceptual model. We also proposed a SoS reliability definition grounded on the literature. By organizing the knowledge of SoS reliability, this conceptual model makes it possible to expand the body of knowledge in the area and opens several opportunities for further investigations; in particular, this model serves as a basis for novel solutions aiming to assure SoS reliability.

Phosphorylation of PACSIN2 at S313 Regulates Podocyte Architecture in Coordination with N-WASP.

Changes in the dynamic architecture of podocytes, the glomerular epithelial cells, lead to kidney dysfunction. Previous studies on protein kinase C and casein kinase 2 substrates in neurons 2 (PACSIN2), a known regulator of endocytosis and cytoskeletal organization, reveal a connection between PACSIN2 and kidney pathogenesis. Here, we show that the phosphorylation of PACSIN2 at serine 313 (S313) is increased in the glomeruli of rats with diabetic kidney disease. We found that phosphorylation at S313 is associated with kidney dysfunction and increased free fatty acids rather than with high glucose and diabetes alone. Phosphorylation of PACSIN2 emerged as a dynamic process that fine-tunes cell morphology and cytoskeletal arrangement, in cooperation with the regulator of the actin cytoskeleton, Neural Wiskott-Aldrich syndrome protein (N-WASP). PACSIN2 phosphorylation decreased N-WASP degradation while N-WASP inhibition triggered PACSIN2 phosphorylation at S313. Functionally, pS313-PACSIN2 regulated actin cytoskeleton rearrangement depending on the type of cell injury and the signaling pathways involved. Collectively, this study indicates that N-WASP induces phosphorylation of PACSIN2 at S313, which serves as a mechanism whereby cells regulate active actin-related processes. The dynamic phosphorylation of S313 is needed to regulate cytoskeletal reorganization.

A Secure Intrusion Detection Platform Using Blockchain and Radial Basis Function Neural Networks for Internet of Drones

The Internet of Drones (IoD) is built on the Internet of Things (IoT) by replacing "Things" with "Drones" while retaining incomparable features. Because of its vital applications, IoD technologies have attracted much attention in recent years. Nevertheless, gaining the necessary degree of public acceptability of IoDs without demonstrating safety and security for human life is exceedingly difficult. In addition, Intrusion Detection Systems (IDSs) in IoD confront several obstacles because of the dynamic network architecture, particularly in balancing detection accuracy and efficiency. To increase the performance of the IoD network, we proposed a blockchain-based Radial Basis Function Neural Networks (RBFNN) model in this paper. The proposed method can improve data integrity and storage for smart decision-making across different IoDs. We discussed the usage of blockchain to create decentralized predictive analytics and a model for effectively applying and sharing Deep Learning (DL) methods in a decentralized fashion. We also assessed the model using a variety of datasets to demonstrate the viability and efficacy of implementing the blockchain-based DL technique in IoD contexts. The findings showed that the suggested model is an excellent option for developing classifiers while adhering to the constraints placed by network intrusion detection. Furthermore, the proposed model can outperform the cutting-edge methods in terms of specificity, F1, recall, precision, and accuracy.

Accurate online training of dynamical spiking neural networks through Forward Propagation Through Time

With recent advances in learning algorithms, recurrent networks of spiking neurons are achieving performance that is competitive with vanilla recurrent neural networks. However, these algorithms are limited to small networks of simple spiking neurons and modest-length temporal sequences, as they impose high memory requirements, have difficulty training complex neuron models and are incompatible with online learning. Here, we show how the recently developed Forward-Propagation Through Time (FPTT) learning combined with novel liquid time-constant spiking neurons resolves these limitations. Applying FPTT to networks of such complex spiking neurons, we demonstrate online learning of exceedingly long sequences while outperforming current online methods and approaching or outperforming offline methods on temporal classification tasks. The efficiency and robustness of FPTT enable us to directly train a deep and performant spiking neural network for joint object localization and recognition, demonstrating the ability to train large-scale dynamic and complex spiking neural network architectures. Memory efficient online training of recurrent spiking neural networks without compromising accuracy is an open challenge in neuromorphic computing. Yin and colleagues demonstrate that training a recurrent neural network consisting of so-called liquid time-constant spiking neurons using an algorithm called Forward-Propagation Through Time allows for online learning and state-of-the-art performance at a reduced computational cost compared with existing approaches.

An SSL-PUF Based Access Authentication and Key Distribution Scheme for the Space-Air-Ground Integrated Network.

The Space-Air-Ground Integrated Network (SAGIN) expands cyberspace greatly. Dynamic network architecture, complex communication links, limited resources, and diverse environments make SAGIN's authentication and key distribution much more difficult. Public key cryptography is a better choice for terminals to access SAGIN dynamically, but it is time-consuming. The semiconductor superlattice (SSL) is a strong Physical Unclonable Function (PUF) to be the hardware root of security, and the matched SSL pairs can achieve full entropy key distribution through an insecure public channel. Thus, an access authentication and key distribution scheme is proposed. The inherent security of SSL makes the authentication and key distribution spontaneously achieved without a key management burden and solves the assumption that excellent performance is based on pre-shared symmetric keys. The proposed scheme achieves the intended authentication, confidentiality, integrity, and forward security, which can defend against masquerade attacks, replay attacks, and man-in-the-middle attacks. The formal security analysis substantiates the security goal. The performance evaluation results confirm that the proposed protocols have an obvious advantage over the elliptic curve or bilinear pairings-based protocols. Compared with the protocols based on the pre-distributed symmetric key, our scheme shows unconditional security and dynamic key management with the same level performance.

Privacy protection in intelligent vehicle networking: A novel federated learning algorithm based on information fusion

Federated learning is an effective technique to solve the problem of information fusion and information sharing in intelligent vehicle networking. However, most of the existing federated learning algorithms generally have the risk of privacy leakage. To address this security risk, this paper proposes a novel personalized federated learning with privacy preservation (PDP-PFL) algorithm based on information fusion. In the first stage of its execution, the new algorithm achieves personalized privacy protection by grading users’ privacy based on their privacy preferences and adding noise that satisfies their privacy preferences. In the second stage of its execution, PDP-PFL performs collaborative training of deep models among different in-vehicle terminals for personalized learning, using a lightweight dynamic convolutional network architecture without sharing the local data of each terminal. Instead of sharing all the parameters of the model as in standard federated learning, PDP-PFL keeps the last layer local, thus adding another layer of data confidentiality and making it difficult for the adversary to infer the image of the target vehicle terminal. It trains a personalized model for each vehicle terminal by “local fine-tuning”. Based on experiments, it is shown that the accuracy of the proposed new algorithm for PDP-PFL calculation can be comparable to or better than that of the FedAvg algorithm and the FedBN algorithm, while further enhancing the protection of data privacy.

Order-Based Schedule of Dynamic Topology for Recurrent Neural Network

It is well-known that part of the neural networks capacity is determined by their topology and the employed training process. How a neural network should be designed and how it should be updated every time that new data is acquired, is an issue that remains open since it its usually limited to a process of trial and error, based mainly on the experience of the designer. To address this issue, an algorithm that provides plasticity to recurrent neural networks (RNN) applied to time series forecasting is proposed. A decision-making grow and prune paradigm is created, based on the calculation of the data’s order, indicating in which situations during the re-training process (when new data is received), should the network increase or decrease its connections, giving as a result a dynamic architecture that can facilitate the design and implementation of the network, as well as improve its behavior. The proposed algorithm was tested with some time series of the M4 forecasting competition, using Long-Short Term Memory (LSTM) models. Better results were obtained for most of the tests, with new models both larger and smaller than their static versions, showing an average improvement of up to 18%.

Dynamic architecture of multifunctional community centers

Purpose: The aim of this paper is to highlight the main advantages of adaptability of community centers in the conditions of their rapid development to the needs of the population.Methodology: The interdisciplinary analysis of the global architectural practice concerning the architectural situation in the community centres of cities and towns of our country.Research findings: The main findings are presented as possible ways of implementing modern, innovative design solutions based on the Russian and global experience, previously successfully tested and applied in the completed projects and constructed buildings.Originality: Generalized directions of the community centre formation are proposed using the dynamic architecture concept.