Search

Many sensitive applications require that data remain confidential and undisclosed, even for intrusion detection objectives. For this purpose, the detection of anomalies in encrypted data has become increasingly vital. Deep learning models are becoming good tools to detect anomalies in encrypted data without the need to pass through data decryption. This paper presents a systematic review focusing on the advancements made in deep learning models for intrusion detection over encrypted data with privacy preservation. This study aims to guide researchers on how to select the right tools to set up an intrusion detection system over encrypted data with privacy preservation. The study presented the context and challenges of intrusion detection on encrypted data and how machine learning-based solutions can circumvent these challenges. The paper looks at recently proposed solutions, examines metrics for assessing model performance, and evaluates frequently used reference datasets. Deep learning models are also evaluated with statistics on the most frequent models, datasets, and encryption tools. The performance metrics of the studied solutions are investigated as a function of the encryption tools, the deployed deep learning models, the privacy preservation tools, the deployed datasets, and the eventual additional tools and algorithms. Our recommendations help researchers evaluate their proposals for preserving privacy and detecting intrusions on encrypted data using deep learning techniques. [ABSTRACT FROM AUTHOR]

The existing joint extraction methods for entity relationships in knowledge data only target specific fields or datasets, which may have insufficient coverage for large-scale and diverse manufacturing knowledge data. In order to achieve more accurate and efficient joint extraction of manufacturing knowledge data entity relationship, an improved method based on BERT algorithm is proposed in this paper. This method constructs a manufacturing specific knowledge graph, treating the extraction of quantitative knowledge as the extraction of manufacturing entity attributes, thereby achieving joint extraction of knowledge reasoning and data entity relationships. The improved BERT algorithm was used to establish a distribution feature set of entity relationships in manufacturing knowledge data, and a manufacturing specific knowledge graph was constructed, providing an effective way to analyze and extract entity relationships in manufacturing data, thereby helping decision-makers better understand knowledge in the manufacturing field and improving production efficiency and quality. The test results show that the knowledge graph generated by the method has good expression ability and strong logical reasoning ability, and the running time required by the application of the method is only 12 s, indicating that it can realize the joint clustering extraction of entity relationships in manufacturing knowledge data more efficiently and accurately. The innovation of this paper lies in applying BERT algorithm to joint extraction of entity relationship of manufacturing knowledge data, and improving the algorithm to improve the accuracy and efficiency of extraction, providing a new solution for the construction and application of manufacturing knowledge graph. [ABSTRACT FROM AUTHOR]

Disjoint paths are defined as paths between the source and destination nodes where the intermediate nodes in any two paths are disjoint. They are helpful in fault-tolerance routing and securing message distribution in the network. Several research papers were proposed to solve the problem of finding disjoint paths for a variety of interconnection networks such as Hypercube, Generalized Hypercube, Mesh, Torus, Gaussian, Eisenstein–Jacobi, and many other topologies. In this research, we have developed a general algorithm that constructs maximal node-to-node disjoint paths for symmetric networks where all paths are shortest. The algorithm presented in this paper outperforms other algorithms in finding not only the disjoint paths but shortest and maximal disjoint paths with a complexity of O (n 2) . In addition, we have simulated the proposed algorithm on different networks. The solution of unsolved problem in Cube-Connected-Cycles is given in the simulation results. [ABSTRACT FROM AUTHOR]

Recently, many image encryption algorithms have been proposed. Part of them failed to pass the cryptanalysis test and were found to be unsecure. Some others were more secure and could resist the known cryptanalysis methods. Many of those recent schemes followed a well-known model that consists of encrypting the image's pixels using a variety of techniques and methods. Many algorithms were claimed to be secure and performant to be used in real applications. This paper investigates this claim by performing a comparative analysis between a chaotic algorithm recently proposed and the standard general purpose AES cryptosystem in a web application context. The experimental tests conducted on both algorithms in this paper proved that the pixel-wise algorithms are not suitable for this kind of applications. [ABSTRACT FROM AUTHOR]

This paper proposes a novel hybrid metaheuristic algorithm called the remora crayfish optimization algorithm (HRCOA) designed for solving continuous optimization problems. The crayfish optimization algorithm (COA), recently proposed as a meta-heuristic algorithm (MA), exhibits certain limitations such as imbalanced exploration and exploitation capacities, susceptibility to premature optimization, and a propensity for stagnation. To address these shortcomings, we incorporate the exploitation operators from the remora optimization algorithm (ROA) to enhance the exploitative behaviors of COA. In addition, we simplify the summer resort operator in the original COA to streamline the search operator design, thus avoiding unnecessary complexity. Furthermore, numerical experiments on 10-dimensional (D) and 20-D CEC2022 benchmark functions, 50-D and 100-D CEC2020 benchmark functions, engineering optimization problems, and wireless sensor networks (WSNs) coverage optimization problems are conducted to investigate the performance of our proposed HRCOA comprehensively. We compare the proposed HRCOA against eight well-known state-of-the-art MAs, including CMAES and the original COA, as competitor algorithms. The experimental and statistical results confirm the effectiveness, competitiveness, and scalability of our proposal. Finally, we conclude that the proposed HRCOA possesses significant potential for addressing diverse optimization challenges in real-world scenarios. [ABSTRACT FROM AUTHOR]

By decoupling the data plane and the control plane of networking devices, Software Defined Network (SDN) enables the control plane devices to obtain the abstract view of the entire network and accomplish network functions as necessary. In SDN, control plane device, known as SDN controller is responsible for managing and setting rules for the data plane devices (SDN switches). Multiple SDN controllers are often used to formulate distributed controller architecture for geographically larger networks where different sets of switches are connected to and managed by different controllers. Due to network dynamics, any particular controller can become overloaded which results in increased response time and thereby reduce the end-user quality of service. To tackle the controller overload issue, in this paper, an effective and efficient load balancing framework is proposed. Most of the previous works in this domain focus on precise load balancing; therefore many load shifting (switch migration) operations are required and consequently degrade the end-user QoS. Additionally, beneath a certain load, CPU performs almost equally well; therefore load balancing before reaching a certain controller's load seems unnecessary and questionable. Unlike most of the previous works, our algorithm balances the load loosely in a proper distributed controller SDN architecture. The load balancing framework uses five key modules to perform faster switch migrations to balance the load. Number of switch migrations, selection time performance, and load balancing precision are the performance metrics for evaluation and comparison for this work. The resultant load-balancing solution has successfully shown its potency in terms of reducing the number of switch migrations and the selection time compared to some recent and notable works in this domain. Moreover, a completely new functionality named Switch Restoration is introduced in our load balancing solution to optimize the network further. [ABSTRACT FROM AUTHOR]

This paper addresses the critical gap in the understanding of the effects of various configurations and feedback mechanisms on the performance of hybrid metaheuristics (HMs) in unsupervised clustering applications. Despite the widespread use of HMs due to their ability to leverage multiple optimization methods, the lack of comprehensive studies on their configuration and feedback mechanisms effects often results in sub-optimal clustering performances and premature convergence. To tackle these issues, we introduce two algorithms for implementing eight distinct HM schemes, focusing on the impacts of parallel and serial processing models along with different feedback mechanisms. Our approach involves selecting candidate metaheuristics based on a mix of evolutionary and swarm-based methods, including the k-means algorithm, to form various HM-based clustering schemes. These schemes were then rigorously evaluated across a range of datasets and feedback mechanisms, further assessing their efficiency in the deployment of smart grid base stations. Performance analysis was based on total fitness evaluations, timing capabilities, and clustering accuracy. The results revealed that parallel HMs with decoupled feedback mechanisms performed best in terms of accuracy but at the cost of slower convergence rates as compared to serial HMs. Our findings further suggest that serial HMs will be best suited for time-sensitive applications where a compromise between speed and accuracy is acceptable, while parallel HMs with decoupled feedback mechanisms are preferable for scenarios where precision is paramount. This research significantly contributes to the field by providing a detailed analysis of HM performance in varying conditions, thereby guiding the selection of appropriate HM schemes for specific clustering tasks. [ABSTRACT FROM AUTHOR]

This paper introduces an innovative method to improve loop closure detection within the domain of Simultaneous Localization And Mapping (SLAM) by integrating a Multi-Scale Attention and Semantic Guidance (MSA-SG) framework. In SLAM systems, accurate loop closure detection is essential for minimizing localization errors over time and ensuring the reliability of the constructed maps in robotics navigation through uncharted environments. Our proposed method utilizes EfficientNet-EA for robust feature extraction and introduces MSA-SG, a novel mechanism that synergizes multiscale attention with semantic guidance to focus on critical semantic features essential for loop closure detection. This approach ensures the prioritization of static environmental landmarks over transient and irrelevant objects, significantly enhancing the accuracy and efficiency of loop closure detection in complex and dynamic settings. Experimental validations on recognized datasets underscore the superiority of our approach, demonstrating marked improvements in precision, recall, and overall SLAM performance. This research highlights the significant benefits of leveraging semantic insights and attentional focus in advancing the capabilities of loop closure detection for SLAM applications. [ABSTRACT FROM AUTHOR]

Cuckoo search (CS) algorithm is a classical swarm intelligence algorithm widely used in a variety of engineering optimization problems. However, its search accuracy and convergence speed still have a lot of room for improvement. In this paper, an improved version of the CS algorithm based on intelligent perception strategy, adaptive invasive weed optimization (AIWO), and elite cross strategy, called IIC-CS is proposed. Firstly, the intelligent perception strategy can update the value according to the searching state. Moreover, the CS is hybridized with the AIWO to improve the searching performance of the algorithm. Additionally, the elite cross strategy is employed to enhance the exploration capability and exploitation capability of the algorithm. Combining the improvements of these three methods, the performance of the CS algorithm is significantly improved. Meanwhile, 23 classical benchmark functions, some CEC2014 and CEC2018 benchmark functions are used to test the search accuracy and convergence rate of the IIC-CS. Furthermore, some classical or state-of-the-art algorithms such as the genetic algorithm (GA), particle swarm optimization (PSO), bat algorithm (BA), ant lion optimizer (ALO) and cuckoo search (CS) algorithm, invasive weed optimization (IWO), integrated cuckoo search optimizer (ICSO) and improved island cuckoo search (iCSPM2) are used to make comparisons. Through the statistical results of the experiments, we find that the IIC-CS algorithm can achieve better results on most benchmark functions compared to other algorithms, thus demonstrating the effectiveness of the improvements and the superiority of the IIC-CS algorithm. [ABSTRACT FROM AUTHOR]

Managing the explosion of data from the edge to the cloud requires intelligent supervision, such as fog node deployments, which is an essential task to assess network operability. To ensure network operability, the deployment process must be carried out effectively regarding two main factors: connectivity and coverage. The network connectivity is based on fog node deployment, which determines the network's physical topology, while the coverage determines the network accessibility. Both have a significant impact on network performance and guarantee the network quality of service. Determining an optimum fog node deployment method that minimizes cost, reduces computation and communication overhead, and provides a high degree of network connection coverage is extremely hard. Therefore, maximizing coverage and preserving network connectivity is a non-trivial problem. In this paper, we propose a fog deployment algorithm that can effectively connect the fog nodes and cover all edge devices. Firstly, we formulate fog deployment as an instance of multi-objective optimization problems with a large search space. Then, we leverage Marine Predator Algorithm (MPA) to tackle the deployment problem and prove that MPA is well-suited for fog node deployment due to its rapid convergence and low computational complexity, compared to other population-based algorithms. Finally, we evaluate the proposed algorithm on a different benchmark of generated instances with various fog scenario configurations. Our algorithm outperforms state-of-the-art methods, providing promising results for optimal fog node deployment. It demonstrates a 50% performance improvement compared to other algorithms, aligning with the No Free Lunch Theorem (NFL Theorem) Theorem's assertion that no algorithm has a universal advantage across all problem domains. This underscores the significance of selecting tailored algorithms based on specific problem characteristics. [ABSTRACT FROM AUTHOR]

Anthropogenic hazards are unrelenting threat to lives and property, with human irresponsibility emerging as a leading source of urban as well as industrial fires. The complexity of urban structures and crowded layouts make these kinds of fires more lethal. This paper presents an Artificial Intelligence (AI) based framework designed for smart buildings as a solution to the devastating obstacles caused by fire crises. Our system creates a 3D model of the building using floor plans and the A* algorithm for escape route identification. The proposed framework includes a YOLO-based smart monitoring system for the identification and counting of people caught in a fire, with the ability to distinguish between conscious and unconscious persons. The proposed system informs inhabitants in the case of a fire and directs them to the closest exit for a safe evacuation. Moreover, fire and rescue officials receive real-time information on affected persons, such as the number and location of adults and children who are conscious and unconscious. Perhaps most significantly, the suggested framework performs exceptionally well, scoring 96% for precision and 98% for recall in the detection of fire and humans. These findings highlight the effectiveness of the model in locating people within infrastructures affected by fire. The framework considerably outperforms the most advanced algorithms in terms of speed and efficiency for shortest path detection, greatly improving the ability of fire rescue teams to quickly find and aid residents who are trapped in a fire. [ABSTRACT FROM AUTHOR]

In this paper we propose an auction-based offloading market in the HMC (Heterogeneous Mobile Cloud) environment where mobile devices, Cloudlets, and public cloud form a shared resource pool. In this market, users can register their service offers or submit their offloading requests to the auction manager. Then, the auction manger makes buyer_seller threads as the source–destination of task offloading process. In any made buyer_seller thread a trade-off between conflicting utility functions of buyer and seller has to be guaranteed. Furthermore, auction manager should impose various policies in order to maintain market stability, increase market competition, and reduce monopolies. A stable market is a market where the chance of uncommitted sellers to be considered in the made buyer_seller thread is kept low. Also, almost equal opportunities are made for the users in a competitive and non-monopolized market. To address these issues, first of all, a new task offloading approach based on Hill Climbing Algorithm is designed to assure a trade-off between buyers' and sellers' utilities. Then, a new factor in name reputation is defined to make competitive market while maintain market stability. Furthermore, a Cloudlet selection algorithm is developed to find the Cloudlet as a new offloading destination in a case of reneging on the offloading commitment, leading to market stability improvement. Finally, awarding bonus to new participants is another strategy imposed by auction manager to make competitive and non-monopolized market. Extensive amounts of simulation were carried out to evaluate the performance of our proposed offloading market against (Zhou et al. in J Syst Softw 152:151–164, 2019). The results show that our designed market outperforms (Zhou et al. 2019). [ABSTRACT FROM AUTHOR]

Message Passing Interface (MPI) is the de facto standard for parallel programming, and collective operations in MPI are widely utilized by numerous scientific applications. The efficiency of these collective operations greatly impacts the performance of parallel applications. With the increasing scale and heterogeneity of HPC systems, the network environment has become more complex. The network states vary widely and dynamically between node pairs, and this makes it more difficult to design efficient collective communication algorithms. In this paper, we propose a method to optimize collective operations by using real-time measured network states, specifically focusing on the binomial tree algorithm. Our approach employs a low-overhead method to measure the network states, and the binomial tree with small latency is constructed based on the measurement result. Additionally, we take into account the disparities between the two underlying MPI peer-to-peer communication protocols, eager and rendezvous, and design tailored binomial tree construction algorithms for each protocol. We have implemented hierarchical MPI_Bcast, MPI_Reduce, MPI_Gather and MPI_Scatter, utilizing our network states-aware binomial tree algorithm at the inter-node level. The benchmark results demonstrate that our algorithm effectively enhances performance in small and medium message communication when compared to the default binomial tree algorithm in Open MPI. Specifically, for MPI_Bcast, we observe an average performance improvement of over 15.5% when the message size is less than 64KB. Similarly, for MPI_Reduce, there is an average performance improvement of over 12.1% when the message is below 2KB. In addition, there is an average performance improvement of over 10% for MPI_Gather when the message ranging from 64B to 512B. For MPI_Scatter, our algorithm achieved performance improvement only for certain message sizes. [ABSTRACT FROM AUTHOR]

Multi-objective optimization problems with large-scale variables are the focus and difficulty of current research in the field of multi-objective evolutionary algorithms. In order to cope with the problem of "dimensional catastrophe", this paper proposes a large-scale multi-objective firefly algorithm based on reward and punishment mechanisms and adaptive dimensional reorganization. First, a reward and punishment mechanisms is proposed to make full use of the information of the solution itself, so that the population can break the constraint and rapidly approach the Pareto frontier, effectively improving the convergence of the population; then, an adaptive dimensional reorganization mechanism is proposed to select individuals with large differences according to different iteration periods to interact and learn from each other, effectively improving the diversity of the population. In the experimental part, to verify the effectiveness of the algorithm, the UF series and the variable expanded ZDT series, 15 test problems are selected for simulation experiments and compared with 8 advanced large-scale optimization algorithms, the results show that L-MOFA-RA has the advantages of fast convergence speed and high convergence accuracy, and the comprehensive experimental results show that compared with the comparison algorithm L-MOFA-RA has better optimization performance. The comprehensive experimental results show that L-MOFA-RA has better optimization performance compared to the comparison algorithm. [ABSTRACT FROM AUTHOR]

The concept and data drift problems have received much attention in recent years. This aspect is crucial in many domains exhibiting non-stationary and cyclical patterns affecting their generative processes. Drift detection can be treated as a supervised task, with labeled data constantly used to validate the learned model. From a practical point of view, this is an impractical task because labeling is complex, costly, and time-consuming. On the other hand, unsupervised change detection techniques are cumbersome in applications because they generate many false alarms. The paper presents a new concept drift detection method based on feature analysis. Stream of data carries information about the distribution patterns that reflect different concepts that may be hidden in the data. The essential features are searched and ranked by LASSO. The rank of features and statistics are employed to feature drift detection. The proposed approach was experimentally checked based on synthetic and natural datasets. The results show that the proposed FBDD algorithm has an advantage over other solutions. [ABSTRACT FROM AUTHOR]

In many fields including aerospace, automotive, and telecommunications, MathWorks' MATLAB/Simulink is contemporary standard for model-based design. The strengths of Simulink are rapid design and algorithm exploration. Models created with Simulink are just functional. Therefore, designers cannot effortlessly consider a Simulink model's architecture. As current architectures are optimized to run on multicore processors, software running on these processors needs to be parallelized in order to benefit from their natural performance. For instance, designers need to understand how a Simulink model could be parallelized and how an adequate multicore architecture is selected. This paper focuses on the dataflow-based parallelization of Simulink models and proposes a method based on dataflow to measure the performance of parallelized Simulink models running on multicore architectures. Throughout the parallelization process, the model is converted into a Hierarchical Synchronous DataFlow Graph (HSDFG) keeping its original semantics, and each composite node in the graph is flattened. Then, the graph is mapped and scheduled into a multicore architecture with the ultimate objective that minimizes the end-to-end latency. In the experiment of applying the proposed approach to a real Simulink model, latency of the parallelized model could be reduced successfully on a various multi-core architectures. [ABSTRACT FROM AUTHOR]

C-RAN (Cloud Radio Access Network) is a 5G architecture that consists of sites and three-layer Data Centers (DCs), which include the central office DC, local DC, and regional DC. Network slicing, which enables infrastructure providers (InP) to create independent logical networks, is essential in this architecture. By utilizing this technology, InPs can maximize the utility of the network by providing slices to service providers in response to their slice requests. However, almost all of the recent research on slice admission control (SAC) schemes has only considered one or two layers of DCs, which limits the efficiency of the slicing process and decreases network utility. To address these issues, this paper proposes an intelligent SAC scheme called ISAC that considers all three-layer DCs. Instead of relying on reinforcement learning algorithms like Q-learning, which are effective in discrete environments with limited state space but give poor performance in continuous environments, ISAC employs the Approximate Reinforcement Learning (ARL) algorithm. ARL is better suited for 5G network modeling because it can adapt to continuous environments, allowing for a more accurate representation of the underlying physical processes. Extensive simulation studies demonstrate that ISAC significantly improves performance in terms of slice request rejection rates, InP revenue, accepting more slices, and optimizing resource utilization. [ABSTRACT FROM AUTHOR]

Offloading computation-intensive tasks of an application from a mobile device to a remote computation resource can improve the runtime of the application as well as improve the lifetime of the mobile device's battery. However, computation offloading is a complex endeavour: remote computation resources in the cloud and in the fog have different computation capabilities and costs, and the network bandwidth and latency between different resources need to be taken into account. In addition, the mobile device would ideally like to offload multiple applications simultaneously, further adding to the complexity of assigning different tasks of each application to available computation resources. In this paper, we introduce an open-source versatile computation offloading simulator (OVCOSim), capable of simulating and optimizing offloading multiple applications to different cloud and fog resources and visualizing the outcome for users. We demonstrate how OVCOSim can be used to optimize assignment of tasks to computation resources to accommodate different user objectives, runtime, battery life, cost, or a combination of these objectives, using a variety of assignment algorithms. We also provide several examples to demonstrate how OVCOSim can be used, as well as validate its functionality. [ABSTRACT FROM AUTHOR]

The synergy of edge computing and Machine Learning (ML) holds immense potential for revolutionizing Internet of Things (IoT) applications, particularly in scenarios characterized by high-speed, continuous data generation. Offline ML algorithms struggle with streaming data as they rely on static datasets for model construction. In contrast, Online Machine Learning (OML) adapts to changing environments by training the model with each new observation in real-time. However, developing OML algorithms introduces complexities such as bias and variance considerations, making the selection of suitable estimators challenging. In this challenging landscape, ensemble learning emerges as a promising approach, offering a strategic framework to navigate the bias-variance tradeoff and enhance prediction accuracy by amalgamating outputs from diverse ML models. This paper introduces a novel ensemble method tailored for edge computing environments, designed to efficiently operate on resource-constrained devices while accommodating various online learning scenarios. The primary objective is to enhance predictive accuracy at the edge, thereby empowering IoT applications with robust decision-making capabilities. Our study addresses the critical challenges of ML in resource-constrained edge computing environments, offering practical insights for enhancing predictive accuracy and scalability in IoT applications. To validate our ensemble's efficacy, we conducted comprehensive experimental evaluations leveraging both synthetic and real-world datasets. The results indicate that our ensemble surpassed state-of-the-art data stream algorithms and ensemble regressors across a range of regression metrics, underlining its superior predictive prowess. Furthermore, we scrutinized the ensemble's performance within the realm of auto-scaling for Virtual Network Function (VNF)-based applications situated at the network's edge, thereby elucidating its applicability and scalability in real-world scenarios. [ABSTRACT FROM AUTHOR]

This paper represents the Reliable Mobility Management of RPL (RM-RPL) protocol, specifically developed to overcome the limitations of the Routing Protocol for Low-Power and Lossy Networks (RPL) in mobile IoT environments. RM-RPL incorporates a sophisticated mechanism to prevent the formation of loops, enabling mobile nodes to operate as both routers and parents within the network. It introduces a novel objective function that optimizes the selection of parent nodes and includes a mechanism to adjust the protocol's behavior when nodes are stationary. Furthermore, an algorithm is devised to acknowledge critical packets properly. The proposed model provides superior support for mobility, efficient routing, and dependable data transmission, rendering it highly suitable for diverse IoT applications. Through comprehensive evaluations, RM-RPL demonstrates exceptional performance in challenging scenarios characterized by large-scale networks, high density, and dynamic conditions. Comparative analysis reveals that RM-RPL significantly enhances the packet delivery ratio and exhibits commendable power consumption, end-to-end delay, and handover delay. [ABSTRACT FROM AUTHOR]

Fog computing has emerged as the most popular technology for processing delay-sensitive tasks in the Internet of Things platform. However, offloading tasks to suitable fog nodes (FNs) for their processing is a very challenging problem which is known to be NP-hard. In this paper, we propose a metaheuristic-based task offloading scheme that jointly optimizes the offloading delay and resource utilization. We first formulate the task offloading as a multi-objective optimization problem and then propose an algorithm based on the Discrete Jaya Algorithm (DJA) to solve it. The DJA is presented with efficient candidate representation and mutation operation. We also propose a task migration algorithm to transfer the partially completed task to another FN before failure. Through extensive simulations, we demonstrate the efficacy of the proposed work by comparing the state-of-the-art algorithms and validating it through the Friedman test. [ABSTRACT FROM AUTHOR]

Differential Evolution (DE) is a population-based metaheuristic search algorithm that exhibits excellent performance. However, it is sensitive to mutation strategies and control parameters.To mitigate the impact of these factors, this paper proposes refining differential evolution with mutation rate and neighborhood weight local search(MRNLDE). The algorithm guides individuals in selecting an appropriate mutation mode through the mutation rate, effectively utilizing the evolutionary information inherent in the individuals themselves, thereby enhancing the search efficiency of the algorithm. Furthermore, the neighborhood individual information is utilized to re-explore the solution space, allowing for a comprehensive exploration of individual evolutionary potential and enhancing the algorithm's diversity. The performance of MRNLDE was validated under two sets of benchmark problems at the Institute of Electrical and Electronics Engineers (IEEE) Conference on Computing in Evolution (CEC), and the results show that MRNLDE performs well overall. [ABSTRACT FROM AUTHOR]

Searching an optimal value of the neural emulator adaptive rate presents a great problem. Indeed, a new scheme of neural emulators based on the Particle Swarm Optimization (PSO) algorithm for nonlinear systems is adopted in this paper. The main goal of this approach consists in adjusting effectively the neural emulator adaptive rate in order to accelerate the convergence speed and to improve the precision degree. The obtained results are compared with those reached with an intelligent tuning strategy. An experimental validation of the new emulator adaptation is carried on chemical reactor. Efficiency of the proposed method is proved according to the obtained performances. [ABSTRACT FROM AUTHOR]

Because of the limitations of current PCB detection methods, such as low detection accuracy and high rate of false and missed detection, a PCB defect detection algorithm based on improved RetinaNet is proposed, that is, RetinaNet-line. Firstly, the algorithm uses PVT v2B2-Hi as the backbone network of the whole model, which can not only solve the deep feature redundancy caused by the residual network in RetinaNet but also improve the feature extraction ability of the network. Then, the CARAFE up-sampling module is used to enrich the semantic features of defective targets and promote the discrimination ability of the algorithm to defective targets. Finally, HSV color space is used to process the input feature image, and the color distribution is expanded by random perturbation in H, S, and V channels, with the data enhancement completed. In this paper, experiments are conducted on the publicly shared PCB defect detection dataset from the Intelligent Open Laboratory of Peking University, and the results show that the improved algorithm achieves 94.2% in the mAP index, which is better than other networks. [ABSTRACT FROM AUTHOR]

The overwhelming growth of cloud computing has introduced new challenges for the cloud infrastructure. Many cloud service providers (CSPs) have been adopting renewable energy (RE) sources to increase profitability and reduce carbon emissions. Therefore, recent literature focuses on managing cloud infrastructure with RE sources in addition to traditional non-renewable energy (NRE) sources whenever required. However, these works aim to maximize the usage of RE or minimize the cost without considering the uncertainty (UN). This paper presents three user request-based scheduling algorithms, namely UN-based future-aware best fit, UN-based round-robin, and UN-based highest available renewable first, by managing the UN of RE and NRE. These algorithms consider two types of UN, namely UN-RE and UN-NRE, concerning the user request. The proposed algorithms undergo a rigorous simulation process with 200 to 2000 user requests and 20 to 200 datacenters and are compared using overall cost, UN cost and time, and the number of used RE resources. [ABSTRACT FROM AUTHOR]

IoT networks can provide many benefits and opportunities, although their implementation poses challenges. Cloud-only storage of IoT data would be very costly and time-consuming. In this paper, a new scheme is proposed for caching IoT content on the edge with SDN-based processing capability. The proposed scheme considers a global SDN controller, which coordinates cache decisions across the entire IoT network. The Moth-Flame Optimization-Cluster Head Selection (MFO-CHS) algorithm is used to cluster devices where the selected cluster heads send the IoT data to the edge nodes for caching. In addition, by utilizing edge caching capabilities and using MFO to select and cache the appropriate contents on edge nodes, the proposed Moth-Flame Optimization-Edge Caching (MFO-EC) algorithm can provide data with lower latency on upcoming requests. Caching can also help ensure reliability and availability since intermittent connections and power limitations affect IoT devices. Caching decisions regarding IoT characteristics are not made intelligently in the default caching scheme for maximizing device longevity and managing the possibility that content producers may become unreachable. This scheme considers several metrics, and the proposed moth-flame optimization (MFO) algorithms, MFO-CHS, and MFO-EC algorithms, which are nature-inspired paradigms for the edge caching problem called "MFO-SDN-EC", Moth-Flame Optimization-Software-defined Networking -Edge Caching, are used to select the best options regarding considered criteria and improve the QoS in the SDN based IoT environment. Based on simulations, our proposed caching method can reduce energy consumption by 45%, decrease the average response time by 49%, and also increase cache-hit rates. Furthermore, our results demonstrate our algorithm's superiority over several current approaches in terms of assessment measures. [ABSTRACT FROM AUTHOR]

Due to easier access, improved performance, and lower costs, the use of cloud services has increased dramatically. However, cloud service providers are still looking for ways to complete users' jobs at a high speed to increase profits and reduce energy consumption costs. To achieve such a goal, many algorithms for scheduling problem have been introduced. However, most techniques consider an objective in the scheduling process. This paper presents a new hybrid multi-objective algorithm, called SMO_ACO, for addressing the scheduling problem. The proposed SMO_ACO algorithm combines Spider Monkey Optimization (SMO) and Ant Colony Optimization (ACO) algorithm. Additionally, a fitness function is formulated to tackle 4 objectives of the scheduling problem. The proposed fitness function considers parameters like schedule length, execution cost, consumed energy, and resource utilization. The proposed algorithm is implemented using the Cloud Sim toolkit and evaluated for different workloads. The performance of the proposed technique is verified using several performance metrics and the results are compared with the most recent existing algorithms. The results prove that the proposed SMO_ACO approach allocates resources efficiently while maintaining cloud performance that increases profits. [ABSTRACT FROM AUTHOR]

The large amount of data generated every day makes necessary the re-implementation of new methods capable of handle with massive data efficiently. This is the case of Association Rules, an unsupervised data mining tool capable of extracting information in the form of IF-THEN patterns. Although several methods have been proposed for the extraction of frequent itemsets (previous phase before mining association rules) in very large databases, the high computational cost and lack of memory remains a major problem to be solved when processing large data. Therefore, the aim of this paper is three fold: (1) to review existent algorithms for frequent itemset and association rule mining, (2)to develop new efficient frequent itemset Big Data algorithms using distributive computation, as well as a new association rule mining algorithm in Spark, and (3) to compare the proposed algorithms with the existent proposals varying the number of transactions and the number of items. To this purpose, we have used the Spark platform which has been demonstrated to outperform existing distributive algorithmic implementations. [ABSTRACT FROM AUTHOR]

Fog Computing paradigm that provisions low-latency computing services at the edge network, is a bonanza for supply chain computing resources in Internet of Things (IoT) applications. In different scenarios such as smart homes/healthcare systems, multiple IoT applications are distributed simultaneously in cloud and fog nodes to provide different IoT-based services. In addition, each program requires resources and has its desired quality of service (QoS) which should be met. One of the key challenges in fog computing environment is how to efficiently allocate resources to IoT applications because inefficient resource allocation leads to burdening providers high costs and it lowers down the delivered QoS to users. Since the majority of IoT applications are time-sensitive, the low delay and near physically allocated resources improve the amount of delivered QoS. Therefore, the resource clustering algorithms with the lowest distance error rate and the lowest delay as a consequence are favorable. The aim is to reduce clustering errors and improve the overall performance of the system. This paper formulates resource allocation to IoT applications in heterogeneous 4-layered fog platforms to an optimization problem. To solve this problem, a fusion approach incorporating a genetic algorithm (GA) and the k-means clustering approach is presented. Firstly, it utilizes the k-means approach and Jaccard measurement to cluster fog nodes with a minimum clustering rate. Then, the resources of fog clusters are allocated to IoT devices with the minimum error rate by incorporating GA algorithm. This selection of processing nodes in a fog layer helps to minimize latency and allows applications to access resources simultaneously. The simulation results in extensive scenarios prove the superiority of the proposed algorithm against other successful meta-heuristic approaches in terms of the objective function and lowest error/delay rate. [ABSTRACT FROM AUTHOR]

Cloud computing has experienced extraordinary development across a wide range of industries by giving customers the flexibility to employ computing resources as needed. The task scheduling problem is one of several major challenges in cloud computing, and it should be scheduled effectively to minimize makespan and maximize resource utilization. Therefore, this paper put forward an improved scheduling efficiency algorithm called Enhanced Marine Predator Algorithm (EMPA). Firstly, task scheduling model with makespan and resource utilization is constructed. Secondly, each individual represents a result of task scheduling, and the purpose of algorithms is to find the optimal scheduling result, therefore the operator of WOA, nonlinear inertia weight coefficient and golden sine strategy are introduced into Marine Predator Algorithm. In the simulation experiment, EMPA is compared with Grey Wolf Optimizer (GWO), Sine Cosine Algorithm (SCA), Particle Swarm Optimization (PSO), and Whale Optimization Algorithm (WOA) under different number of tasks in synthetic datasets and GoCJ datasets.The experimental results show that the EMPA algorithm has more advantages in terms of makespan, degree of imbalance, and resource utilization. [ABSTRACT FROM AUTHOR]

In this paper, we consider the matter of on-line clustering of hyperplanes within the presence of bounded noise. This is often a challenging problem that involves the segmentation of the data and the estimation of the hyperplanes parameters. The proposed algorithm consists in two successive steps. At whenever, the first step allows to assign the current data point to the most appropriate hyperplane and therefore the second step realizes an update of the parameters of the hyperplane that contains the data point. The second step springs from an Outer Bounding Ellipsoid type algorithm suitable for on-line parameters estimation within the presence of bounded noise. The performance of our algorithm is proven using synthetic and real data. [ABSTRACT FROM AUTHOR]

Numerous data centers have adopted heterogeneous computing server to accelerate the processing speed with various applications. However, as the growth of efficiency, the issue of high power consumption is increasingly becoming crucial. Traditional researchers focus on the developing strategies for power efficiency that neglects the importance of power prediction that substantially impacts on energy planning on data center. Different to the previous work, this paper utilizes the advantage of Artificial Neural Network (ANN) to design and implement an improved algorithm of power consumption prediction on heterogeneous computing server with various types of jobs. First, the impact of multi-type of jobs has been taken into account, i.e., CPU-intensive, memory-intensive, I/O-intensive and GPU-intensive. We collect the data trace from these jobs by a set of benchmarks. Then, based on the collected data trace, power prediction algorithm has been established, so that developing an improved ANN-based power consumption prediction algorithm with CNN-BiLSTM neural network and attention mechanism. Specifically, one dimensional CNN is utilized to perform local feature extraction and dimension reduction on the input, the BiLSTM network is adopted to extract high-level features from the above results. The miscellaneous features extracted are filtered out by using attention mechanism. Last, the experiment has been conducted, and the peculiarity for four types of jobs has been discussed. Compared with the latest proposed ANN-based prediction power algorithms, our proposed algorithm has better performance in prediction accuracy on power consumption. [ABSTRACT FROM AUTHOR]

Clustering is an important technique for data mining and machine learning. In this paper, a novel clustering algorithm, scale division and region growing (SRG), is proposed, which is based on scale division and region growing. This method uses the division-and-merge strategy, which consists of two primary phases, namely the division phase and the merging phase. In the division phases, the data is mapped to a two-dimensional space, and the clustering problem is transformed to an image processing problem. Then the image is divided into different scales to adaptively find the appropriate scale. After finding the appropriate scale, the region growing is carried out under this scale and the multiple sub-clusters are created. In the merging phases, the sub-clusters are evaluated and merged to form the final clusters in the dataset. The proposed algorithm is tested on 15 benchmark datasets including both the synthetic datasets and the real datasets, and it is found that SRG can obtain better clustering results and has higher accuracy than other typical clustering algorithms. Moreover, the experimental results on noisy datasets show that the proposed algorithm has good robustness. [ABSTRACT FROM AUTHOR]

As the performance of Radio Frequency Identification (RFID) devices is susceptible to the influence of the surrounding environment, making the original data collected by RFID devices with uncertain, of which missed data and redundant data are the main source of uncertainty data, these uncertainty data will seriously affect the quality of RFID upper layer applications. Therefore, to solve the uncertainty in RFID data, the original data must be cleaned. For the shortcomings of tag dynamics detection in the traditional data cleaning algorithm named Statistical Smoothing for Unreliable RFID data (SMURF), an adaptive sliding window-based data cleaning algorithm for RFID single-tag is proposed. The method takes into account the influence of tag speed on tag integrity judgment, and divides sliding sub-windows to accurately detect tag changes and then reasonably adjusts the sliding window size. In addition, considering that the average reading rate of tags is affected by the collision between multiple tags, an RFID multi-tag cleaning method based on twice-tag number estimation is proposed. The method accurately estimates the number of tags by the twice-tag estimation method, controls the read period by the Markov chain, and reduces the occurrence of multiple tag collisions by using an unequal time slot optimization method. Experimental results show that the proposed method in this paper can form a complete set of RFID data stream cleaning algorithms, which effectively reduces the uncertain data and improves the data accuracy. [ABSTRACT FROM AUTHOR]