Search

Controlling complex dynamical systems has been a topic of considerable interest in academic circles in recent decades. While existing works have primarily focused on closed-loop control schemes with infinite-time durations, this paper introduces a novel finite-time, closed-loop stochastic controller that pays special attention to control time and energy and their dependence on system parameters. This technique of stochastic control not only enables finite-time control in chaotic dynamical systems but also facilitates finite-time synchronization in unidirectionally coupled systems. Notably, our new scheme offers several advantages over existing deterministic finite-time controllers from a physical standpoint of time and energy consumption. Using numerical experiments based on random ecosystems, neural networks, and Lorenz systems, we demonstrate the effectiveness of our analytical results. It is anticipated that this proposed stochastic scheme will have widespread applicability in controlling complex dynamical systems and achieving network synchronization., Comment: There are some mistakes in the paper. We need to revise it

Hierarchical temporal memory is an emerging machine learning technology that aims to model the structural and algorithmic properties of the neocortex. It is particularly suitable for learning and p...

VANET is the spontaneous evolving creation of a wireless network, and clustering in these networks is a challenging task due to rapidly changing topology and frequent disconnection in networks. The cluster head (CH) stability plays a prominent role in robustness and scalability in the network. The stable CH ensures minimum intra- and intercluster communication, thereby reducing the overhead. These challenges lead the authors to search for a CH selection method based on a weighted amalgamation of four metrics: befit factor, community neighborhood, eccentricity, and trust. The stability of CH depends on the vehicle’s speed, distance, velocity, and change in acceleration. These all are included in the befit factor. Also, the accurate location of the vehicle in changing the model is very vital. Thus, the predicted location with the Kalman filter’s help is used to evaluate CH stability. The results have shown better performance than the existing state of the art for the befit factor. The change in dynamics and frequent disconnection in communication links due to the vehicle’s high speed are inevitable. To comprehend this problem, a graphing approach is used to evaluate the eccentricity and the community neighborhood. The link reliability is calculated using the eigengap heuristic. The last metric is trust; this is one of the concepts that has not been included in the weighted approach to date as per the literature. An adaptive spectrum sensing is designed for evaluating the trust values specifically for the primary users. A deep recurrent learning network, commonly known as long short-term memory (LSTM), is trained for the probability of detection with various signals and noise conditions. The false rate has drastically reduced with the usage of LSTM. The proposed scheme is tested on the real map of Chengdu, southwestern China’s Sichuan province, with different vehicular mobilities. The comparative study with the individual and weighted metric has shown significant improvement in the cluster head stability during high vehicular density. Also, there is a considerable increase in network performance in energy, packet delay, packet delay ratio, and throughput.

The need to capture intra-word information in natural language processing (NLP) tasks has inspired research in learning various word representations at word, character, or morpheme levels, but little attention has been given to syllables from a syllabic alphabet. Motivated by the success of compositional models in morphological languages, we present a Convolutional-long short term memory (Conv-LSTM) model for constructing Swahili word representation vectors from syllables. The unified architecture addresses the word agglutination and polysemous nature of Swahili by extracting high-level syllable features using a convolutional neural network (CNN) and then composes quality word embeddings with a long short term memory (LSTM). The word embeddings are then validated using a syllable-aware language model ( 31.267 ) and a part-of-speech (POS) tagging task ( 98.78 ), both yielding very competitive results to the state-of-art models in their respective domains. We further validate the language model using Xhosa and Shona, which are syllabic-based languages. The novelty of the study is in its capability to construct quality word embeddings from syllables using a hybrid model that does not use max-over-pool common in CNN and then the exploitation of these embeddings in POS tagging. Therefore, the study plays a crucial role in the processing of agglutinative and syllabic-based languages by contributing quality word embeddings from syllable embeddings, a robust Conv–LSTM model that learns syllables for not only language modeling and POS tagging, but also for other downstream NLP tasks.

Link prediction is an important task that has wide applications in various domains. However, the majority of existing link prediction approaches assume the given graph follows homophily assumption, and designs similarity-based heuristics or representation learning approaches to predict links. However, many real-world graphs are heterophilic graphs, where the homophily assumption does not hold, which challenges existing link prediction methods. Generally, in heterophilic graphs, there are many latent factors causing the link formation, and two linked nodes tend to be similar in one or two factors but might be dissimilar in other factors, leading to low overall similarity. Thus, one way is to learn disentangled representation for each node with each vector capturing the latent representation of a node on one factor, which paves a way to model the link formation in heterophilic graphs, resulting in better node representation learning and link prediction performance. However, the work on this is rather limited. Therefore, in this paper, we study a novel problem of exploring disentangled representation learning for link prediction on heterophilic graphs. We propose a novel framework DisenLink which can learn disentangled representations by modeling the link formation and perform factor-aware message-passing to facilitate link prediction. Extensive experiments on 13 real-world datasets demonstrate the effectiveness of DisenLink for link prediction on both heterophilic and hemophiliac graphs. Our codes are available at https://github.com/sjz5202/DisenLink

Space information networks is network systems that can receive, transmit, and process spatial information lively. It uses satellites, stratosphere airships, Unmanned Aerial Vehicles, and other platforms as the carrier. It supports high-dynamic, real-time broadband transmission of earth observations and ultra-long-distance, long-delay reliable transmission of deep space exploration. The deeper the network integration, the higher the system’s security concerns and the more likely SINs will be controlled and destroyed in terms of cybersecurity. How to integrate new IT technologies such as artificial intelligence, digital twins, and blockchain to diverse application scenarios of SINs while maintaining SIN cybersecurity will be a long-term critical technical issue. This study is a review of the security issues for space information networks. First, this paper examines space information networks’ security issues and figures out the relationship between the main security threats, services, and mechanisms. Then, this article selects secure routing and anomaly detection from many security technologies to conduct a detailed overview from two perspectives of traditional methods and artificial intelligence. Subsequently, this paper investigates anomaly detection schemes for spatial information networks and proposes a deep learning-based anomaly detection scheme. Finally, we suggest the potential research directions and opening problems of space information network security. Overall, this paper aims to give readers an overview of the newly emerging technologies in space information networks’ security issues and provide inspiration for future exploration.

Training of large scale deep neural networks with distributed implementations is an effective way to improve the efficiency. However, high network communication cost for synchronizing gradients and parameters is a major bottleneck in distributed training. In this work, we propose an asynchronous training algorithm based on sampling and mean fusion for distributed recurrent neural network (RNN). In distributed RNN, multiple distributed neuron nodes and an interaction node work together to implement the training. The synchronization overhead is reduced by a unique asynchronous sampling strategy amongst the distributed neuron nodes. Then, in order to make up for the accuracy loss caused by the asynchronous parameter update, a mean fusion algorithm is proposed, where the interaction node averages all local parameters from the distributed neurons. We mathematically prove the convergence of the proposed algorithm. Experimental verification is performed on two language modeling benchmark datasets. The results demonstrate significant speed gains for distributed RNN, while the accuracy loss is less than 1% on average.

Magnesium is a biocompatible and biodegradable metal, which has attracted much interest in biomedical engineering. Cast magnesium alloy shows the low strength and plasticity at ambient temperature. Microstructure, mechanical properties and degradation properties of the extrusion pressed magnesium alloy have been investigated for biomedical application in detail by optical microscopes, mechanical properties testing and corrosion testing. The magnesium alloy ingots were gained by different cooling rate. Then the ingots were extrude into bar at the same processing condition. The results show that the cooling rate of cast ingot is important factors that affect the properties of Mg alloy by dynamic recrystallisation extruding. The cooling rate of cast ingot has been successfully applied to control the microstructure, mechanical and degradation properties of the Mg alloy. Optical microscopy observation has indicated that the grain size of the dynamic recrystallisation extruding has been significantly decreased from fast cooling cast magnesium ingot, which has mainly contributed to the high tensile strength and good elongation. Fasting cooling rate of cast ingot and dynamic recrystallisation extruding has provided moderate corrosion resistance, which has opened a new window for materials design, especially for biomedical.

Currently, there is a lot of emphasis on emerging engineering education. Under this background, we present a novel challenging experimental course, Information Security System R & D. Specifically, in this paper, we discuss the related course construction background, propose the teaching system and implementation plan of this course, and analyze its effectiveness and innovation. Based on subject competitions, the teaching philosophy of this course embodies student-centric, interest-driven, and project-oriented learning. The proposed course builds a practical teaching platform for the development of information security systems, which provides a valuable reference for the construction of cross-disciplinary teaching platforms in the context of emerging engineering education. More important, the proposed course gradually guides students in developing a practical, innovative, and complete information security system, which explores the potential of implementing “the plan for educating and training outstanding engineers 2.0”.

Development of Internet of Things (IoT) enables smart city advancement throughout the world. Increasing number of vehicles has brought focus on road safety precautions and in-vehicle communication. This is the right time to focus on the development of new applications and services for vehicular environments. The Vehicular Ad-hoc Networks (VANETs) are an interesting range of Mobile Ad-hoc Networks (MANETs) where the Vehicle to Vehicle (V2V) and vehicle roadways transmission is possible. The V2V scheme is fresh by combining Wireless Fidelity (Wi-Fi), Bluetooth and other all sorts of communication standards. An immense number of nodes working with these networks and due to their immense displacements, the analysis is prevailing regarding the possibility of routing standards. The estimation of conventional routing standards for MANETs illustrates that their behaviors are minimal in VANETs. The intention is to make use of mediators for routing with an effort to address the before described issues. The mediators are accountable for gathering data related to routing and identifying the optimal paths for forwarding information packets. The routing scheme is based on group routing standards and data cluster framework for locating the best possible routes. In this paper, we analyze smart cities vehicle communication development by implementing IoT. We also discuss the ways to minimize the limitations connected to IoT deployment and implementation in smart city environment using multi mediator scheme.

As data is growing in different dimensions, it is difficult to get appropriate data analytic tools. Spark is one of high speed "in-memory computing" big data analytic tool designed to improve the efficiency of data computing in both batch and realtime data analytic. Spark is memory bottleneck problem which degrades the performance of applications due to in memory computation and uses of storing intermediate and output result in memory. Investigating how performance is increased in relation to spark executor memory, number of executors, number of cores, and deploy mode parameters configuration in a standalone cluster model is our primary goal. Three representative spark applications are used as workloads to evaluates performance in relation to changing these parameters value. Experimental result show, submitting the job in cluster deploy mode is faster to finish than a submitting job in client deploy mode under two workloads. This implies spark performance does not depend on deploy mode rather it depends on types of application. However, increasing number of executor per worker, a number of core per executor and memory fraction will increase spark performance under all workloads in any deploy mode.

Apache spark is one of the high speed "in-memory computing" that run over the JVM. Due to increasing data in volume, it needs performance optimization mechanism that requires management of JVM heap space. To Manage JVM heap space it needs management of garbage collector pause time that affects application performance. There are different parameters to pass to spark to control JVM heap space and GC time overhead to increase application performance. Passing appropriate heap size with appropriate types of GC as a parameter is one of performance optimization which is known as Spark Garbage collection tuning. To reduce GC overhead, an experiment was done by adjusting certain parameters for loading and dataframe creation and data retrieval process. The result shows 3.23% improvement in Latency and 1.62% improvement in Throughput as compared to default parameter configuration in garbage collection tuning approach.

The development of Location Base Services (LBS) poses new challenges for the protection of user's privacy. Users have to send their current location information to the service provider. The current location of the user can expose critical information such as home/work address, other sensitive information, etc. thus, it is important to protect the users private and sensitive information. As a solution to this problem, different techniques have been proposed over the past few years. Among them, one of the most widely used technique is dummy location generation. However, current dummy location generation methods hardly account for the factor that an attacker has prior knowledge and spatiotemporal information about the user. In this paper, we identify shortcomings and vulnerabilities of the existing techniques and provide a robust solution to maintain the user's data privacy. Furthermore, we propose a robust dummy location generation method capable of averting the negative effects of an attacker having prior knowledge and spatiotemporal information. Additionally, we present some attacker strategies and remedies to avert such attacks. Experiment results show that our proposed method successfully preserves the user's private information, where other state of the art techniques fails to do so.

With the development of petroleum industry, it needs an efficient drill method such as under balanced drilling (UBD) to enhance the rate of penetration (ROP). However, borehole instability is a problem that UBD must face. The coiled tubing partial underbalanced drilling (CT-PUBD) is proposed to try to solve this problem while keeping an underbalanced condition with high ROP. This paper analyzes the laws of cuttings transport in the narrow annulus focus on this new technique through the simulations and experiments. From the results of simulations, it obtains that the particle velocity declines with the increase of rotational speed and increases with the increase of flow rate. The particles become concentrated as the flow rate increases, and the high flow rate limits particles in a small area. The particle distribution undergoes a process of concentration, dispersion, and concentration as the rotational speed increases. The high rotational speed makes particles deviate from the high fluid velocity area, which causes low particle velocity. The relationships between particle velocity and rotational speed and between particle velocity and flow rate are fitted through the equations, respectively. The phenomenon of collision of particles, sinking and rising of particles, and variation of particle velocity are observed in the experiments. The error between the particle velocity in the experiment and numerical simulation is less than 8.5%. This paper is an exploratory study conducted for the cuttings transport in narrow annulus.

Long short-term memory (LSTM) is an important model for sequential data processing. However, large amounts of matrix computations in LSTM unit seriously aggravate the training when model grows larger and deeper as well as more data become available. In this work, we propose an efficient distributed duration-aware LSTM(D-LSTM) for large scale sequential data analysis. We improve LSTM’s training performance from two aspects. First, the duration of sequence item is explored in order to design a computationally efficient cell, called duration-aware LSTM(D-LSTM) unit. With an additional mask gate, the D-LSTM cell is able to perceive the duration of sequence item and adopt an adaptive memory update accordingly. Secondly, on the basis of D-LSTM unit, a novel distributed training algorithm is proposed, where D-LSTM network is divided logically and multiple distributed neurons are introduced to perform the easier and concurrent linear calculations in parallel. Different from the physical division in model parallelism, the logical split based on hidden neurons can greatly reduce the communication overhead which is a major bottleneck in distributed training. We evaluate the effectiveness of the proposed method on two video datasets. The experimental results shown our distributed D-LSTM greatly reduces the training time and can improve the training efficiency for large scale sequence analysis.

Catch statistics are perhaps the most commonly collected data and are widely available for many fisheries. However, it is currently difficult to provide scientific advice for management purposes using only catch data. This article presents a catch-only method for stock assessment of data-poor fisheries. It uses time series of catches and two priors, one for the intrinsic population growth rate derived from life history parameters, and another for stock depletion based on catch trends. The method applies an optimization algorithm to search the potential parameter space. All computations are model or equation based rather than using predefined rules. The utility of this method is demonstrated by applying it to 13 stocks in Australia that are assessed using Stock Synthesis—an assessment package that can make use of a variety of data sources. The estimated parameters, including carrying capacity, intrinsic population growth rate, maximum sustainable yield, and depletion from the catch-only method are broadly comparable with those from the full assessments. The circumstances in which the method may perform poorly, such as longer-term changes in productivity and episodic recruitment, are highlighted.

a b s t r a c t Creep-age forming (CAF) is one of the relatively new forming techniques that has been proven viable for the production of extra-large integral airframe structures. However, experimental studies on forming stiffened structures under creep-ageing conditions remain scarce. In this work, 200 mm × 48 mm inte- grally stiffened plates of aluminium alloy 2219 have been formed on an end clamp device that has a bending radius of 156 mm and creep-aged at 175 ◦C for 18 h. Three different stiffener designs, namely the beam stiffened, waffle and isogrid plates, are tested alongside the flat plates. Utilisation of PTFE pocket fillers and intermediate sheets to reinforce and protect stiffeners during forming operation is found effec- tive as demonstrated by the defect-free formed parts with smooth curvature. Springback of the plates ranged from 12.2 to 15.7% in the experimental studies for different stiffener designs. Using the CAF mate- rial constants determined for this alloy, corresponding finite element models have been developed and experimentally validated using the measured profiles of the creep-age formed plates. Up to 6.3% differ- ence in springback is observed when forming workpieces with different pre-form age condition alone. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND

DC filtering is commonly used together with Fourier algorithm in the calculation of root-mean-square (RMS) values of fault current of power system, so as to eliminate the errors produced by the decaying DC component in the fault current. In this paper, theoretical derivation was conducted, showing that errors of sampling values could be amplified by the DC filtering itself, which could lower the accuracy. So the effect of DC filtering should be comprehensively considered when applied in such circumstance. Numerical simulations concerning different number of sampling points and different initial phase angle of the fault current were also conducted to verify the conclusion. Results indicated that within a certain range of the initial phase angle, the amplification of the sampling error is maximized by the DC filtering, weakening the improvement in accuracy brought by the elimination of DC component, or even worsen the accuracy with too few sampling points for DC filtering calculation.

Póster presentado en la 3rd International Conference on Microbial Diversity (Microbial Diversity 2015. The challenge of Complexity), celebrada en Perugia del 27 al 29 de octubre de 2015., Saccharomyces cerevisiae has an extensive repertoire of molecular methods for the genetic analysis of laboratory strains. Although this has widely contributed to the interpretation of gene functionality within haploid genomes, the genetic basis of desirable metabolic pathways in industrial diploid and polyploid yeast strain is still poorly understood. The genetic engineering of industrial yeasts is currently undergoing major changes due to the development of a marker-free, high-throughput, and multiplexed genome editing approach inspired to the bacterial immune systems: the “Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR-associated (Cas) immune system”. CRISPR/Cas9 allows the knockout or the tuning of expression of specific targeted genes and pathways. The system requires an endonuclease (Cas9) that is guided to a specific DNA sequence by a guide RNA (gRNA). The assembling of a ribonucleoprotein complex (Cas9/gRNA) can produce cuts in specific sites of eukaryotic genomes and generates different combinations of gene deletions through a single transformation event. CRISPR/Cas9 is gaining significant attention because of its biotechnological applications and its potential to boost innovation. The wine industry could particularly gain an advantage from this engineering system; indeed, molecular studies should help understanding the contribution of Saccharomyces and non-Saccharomyces species to several wine features such as those linked to wine quality and safety (aroma and off-flavours compounds, ethanol and glycerol production, sulphur dioxide resistance, toxic compound formation, etc.). In this study we outline a strategy to modify wine yeasts by CRISPR/Cas9 approach. To prove the robustness of this system and to offers an engineering pipeline for further gene editing in specific metabolic pathways relevant for the wine production, two commercial strains of S. cerevisiae (EC1118 and AWRI 796) have been genetically engineered in the arginine degradation pathway to generate strains with a reduced urea and ethyl-carbamate production.

The delay difference model was implemented to fit 21 years of brown tiger prawn (Penaeus esculentus) catch in Moreton Bay by maximum likelihood to assess the status of this stock. Monte Carlo simulations testing of the stock assessment software coded in C++ showed that the model could estimate simultaneously natural mortality in addition to catchability, recruitment and initial biomasses. Applied to logbooks data collected from 1990 to 2010, this implementation of the delay difference provided for the first time an estimate of natural mortality for brown tiger prawn in Moreton Bay, equal to $0.031 \pm 0.002$ week$^{-1}$. This estimate is approximately 30\% lower than the value of natural mortality (0.045 week$^{-1}$) used in previous stock assessments of this species.

Larvae of the golden king crab Lithodes aequispinus were reared in individual chambers in the laboratory from hatching to the first crab stage (C1). The experiment included 2 treatments: 1 group was fed live nauplii of Artemia daily, while the other group was unfed. A total of 4 molts were observed for all larvae from hatching to Cl. One larval stage, either zoea III or zoea IV, was skipped by all larvae. No significant difference existed for the duration of development of each stage between fed and unfed treatments. The mean durations were ZI (ZI = zoeal stage I, ZII = zoeal stage II, etc.) = 6.6 days (±0.71 SD), ZIIa (molted to ZIII) = 7.0 d (±0.65 SD), ZIIb (molted to ZIV) = 8.8 d (±0.95 SD), ZIII (skipped ZIV) = 12.0 d (±1.33 SD), ZIV (skipped ZIII) = 9.3 d (± 1.63 SD), and glaucothoe = 41.3 d (±2.65 SD). The total larval duration was approximately 67 d. The survival rate of the fed group was significantly lower than that of the unfed group. Seventeen percent of unfed larvae developed to Cl, but only 0.7% of fed larvae did so. The longest duration of survival from hatching was 191 d for an unfed crab. The dry weights of unfed larvae decreased 10% from hatching to Cl. Our results indicate that the larval development of the golden king crab is fully lecithotrophic, i.e., larvae can successfully develop to juvenile crabs without eating.

Flooding is widely used in unstructured Peer to Peer (P2P) systems, such as Gnutella. Although it is effective in content search, flooding among peers or super-peers causes large volume of unnecessary traffic. To address this problem, we propose an efficient and adaptive search mechanism, Adaptive Flooding Routing Algorithm (AFRA). AFRA provides the flexibility to adaptively adjust the number of relay neighbors and TTL value to meet different performance requirements. The effectiveness of AFRA is demonstrated through simulation studies. Primary experimental results show that our new AFRA solution reduces about 65% of the flooding messages while maintaining the acceptable high searching quality.

How to construct an interconnected p2p network is an important task in the dynamical environment. This paper presents a new community-based constructing scheme, which can build an interconnected random peer-to-peer network (RP2P) in an efficient and correct way. The basic unit of RP2P is the logical "community", in which one or more peers join in freely. The peers in different communities collaborate to link all the isolated communities into an interconnected network, which in turn allows the peers to directly or indirectly communicate with each other. In RP2P, each peer can enter or leave the communities at will, thus no extra tasks are imposed on the peer. Furthermore, in this paper the RP2P model is also formally described. In order to show how to construct such RP2P, a community based routing algorithm is advanced and analyzed. According to the formula used to compute the connectivity of RP2P, three factors affect its connectivity: the number of the peers, the number of the communities, and the average number of the communities each peer joins. Moreover, this paper argues that, according to the formula, we can improve the connectivity by the means of increasing the number of the peers and the average number of the communities each peer joins. On the contrary, the connectivity decreases with increasing of the communities. The theoretical analysis shows that our novel scheme can provide an interconnected P2P network.

A flexible intrusion detection and response system (ID&R) needs to maximize security while minimizing cost and making response automatically. CI/sup 2/D&R, the cost-based intelligent intrusion detection and response system, is proposed, which is originally developed as a facility to deal with network-based attacks and to make effective response automatically and intelligently. The networking environment deployed with the CI/sup 2/D&R consists of two major parts: guard, which runs on the specific guarded host (GH), and spy, which runs in guarded network (GN). The components of the CI/sup 2/D&R are introduced, which include intrusion detection, attack classification, damage analysis, attack path rebuilding, resources automatically safeguarding, calamity recovery, and security officer. The several kinds of data flow in CI/sup 2/D&R are discussed, too. While CI/sup 2/&R is only a prototype, some experimental results are also presented.

Intrusion detection and response system (ID&R) need to maximize security while minimizing cost and taking response automatically. CI/sup 2/D&R, the Cost-based Intelligent Intrusion Detection and Response System, is originally developed as facilities for deal with network-based attacks automatically and intelligently. This paper provides an overview of CI/sup 2/D&R architecture. The primary components of CI/sup 2/D&R and their functions are also probed into. What is more, the data flow of CI/sup 2/D&R is discussed and compared. Finally, the related work is touched on and conclusions are drawn.

Perovskite solar cells (PSCs) have witnessed astonishing improvement in power conversion efficiency (PCE), more recently, with advances in long-term stability and scalable fabrication. However, the presence of an anomalous hysteresis behavior in the current density-voltage characteristic of these devices remains a key obstacle on the road to commercialization. Herein, sol-gel-processed mesoporous boron-doped TiO2 (B-TiO2) is demonstrated as an improved electron transport layer (ETL) for PSCs for the reduction of hysteresis. The incorporation of boron dopant in TiO2 ETL not only reduces the hysteresis behavior but also improves PCE of the perovskite device. The simultaneous improvements are mainly ascribed to the following two reasons. First, the substitution of under-coordinated titanium atom by boron species effectively passivates oxygen vacancy defects in the TiO2 ETL, leading to increased electron mobility and conductivity, thereby greatly facilitating electron transport. Second, the boron dopant upshifts the conduction band edge of TiO2, resulting in more efficient electron extraction with suppressed charge recombination. Consequently, a methylammonium lead iodide (MAPbI(3)) photovoltaic device based on B-TiO2 ETL achieves a higher efficiency of 20.51% than the 19.06% of the pure TiO2 ETL based device, and the hysteresis is reduced from 0.13% to 0.01% with the B-TiO2 based device showing negligible hysteresis behavior.