Cluster Computing最新文献

The Sparrow Search Algorithm (SSA) is recognized for its rapid convergence and precision in engineering optimization, yet it faces the challenge of premature convergence on complex problems. To address this, a multi-strategy improved sparrow search algorithm (MISSA) is proposed to enhance the optimization performance and applicability in this study. For the first time in the algorithm, the first definition of ellipses is integrated into SSA to balance its exploration and exploitation capabilities. A group co-evolutionary mechanism is introduced to promote population diversity and suppress premature convergence. Unlike most existing work, ablation experiments are utilized to evaluate the effective impact of these enhancement strategies on SSA. Statistical results based on the Wilcoxon signed-rank test and Friedman test show that the dynamic regulator based on the first definition of ellipses has the greatest impact on improving the performance of SSA. Numerical experiments based on the CEC2017 benchmark problems are used as an optimization case to compare MISSA with the classical metaheuristic algorithm and other state-of-the-art variants of SSA. The results demonstrate the outstanding performance and immense potential of MISSA in problem-solving. The applicability of the proposed algorithm is validated through six actual engineering optimization problems, showcasing strong competitiveness in global optimization.

Women are commonly diagnosed with breast cancer (BC), and early detection can significantly increase the cure rate. This study suggested a multi-threshold image segmentation (MTIS) technique for dividing BC histological slice images to assist in identifying lesions and boost diagnostic effectiveness. The selection of the threshold combination, a challenging combinatorial optimization problem, is the key to the MTIS approach. To enhance the MTIS method, a variant of INFO (BQINFO) is proposed to optimize the threshold combination selection procedure. BQINFO is constructed by introducing the barebones mechanism (BM) and quasi-opposition-based learning (QOBL) to INFO and addressing its slow convergence and weakness in local stagnation. To evaluate the optimization performance of BQINFO and the positive impact influence of introducing QOBL and BM to the original INFO for the acceleration of convergence speed and the solution of local stagnation, a series of comparative experiments were carried out using CEC2014 and CEC2021. The comprehensive results and comparisons obtained from the optimization indicators indicate the outstanding performance of BQINFO in overcoming the slow convergence and local stagnation problems when dealing with benchmark function problems. Besides, to further validate BQINFO's performance optimization of threshold combination selection, this paper performed an MTIS experiment with Rényi's entropy as the objective function on BSD500 images and BC histological slice images, respectively, providing qualitative and quantitative analysis with three evaluation metrics, FSIM, PSNR, and SSIM at low and high threshold levels. Ultimately, the experimental results demonstrate that BQINFO performs better and finds the optimal combination of thresholds faster than other comparison algorithms for both low and high threshold levels.

The increasing number of edge layer devices connected to fog servers in fog computing environments has led to a rise in vulnerable and unauthorized actions. Implementing authorized access control with secure key management is essential to address this issue. As the traditional key management methods rely on third-party involvement, which suffers from drawbacks such as single points of failure and inconsistent key management in centralized architecture, so establishing efficient and secure key management between edge devices while ensuring effective access control is the main challenge in the digital environment. This study introduces a novel Blockchain-Based Access Control Protocol in IoT-Enabled Fog Computing (BACP-IeFC) environment for intra-network, inter-network, and mobile device communication models. The BACP-IeFC protocol eliminates the necessity for third-party intermediaries by leveraging Elliptic Curve Cryptography (ECC) for secure data sharing and hash chains for key pair generation. The BACP-IeFC protocol utilizes session keys generated by fog servers, which are securely recorded on a blockchain, ensuring robust authentication at edge devices. A Permissioned Blockchain is also used for secure key storage at the fog layer. The BACP-IeFC security has undergone comprehensive evaluation, including testing its session key (SK) security under the Real-or-Random (ROR) model, confirming its effectiveness in achieving SK security. An informal security analysis confirms the BACP-IeFC protocol resilience against known attacks. For the formal security verification, the BACP-IeFC protocol utilized the ProVerif security tool, and the results show that it is secure against major attacks. Additionally, the performance analysis of the proposed protocol using MIRACL shows a significant improvement in computation overhead, communication, storage cost, and energy consumption cost compared to existing protocols. The scalability and latency analysis of the BACP-IeFC protocol demonstrates that it supports high scalability with low latency costs. The BACP-IeFC protocol is implemented on Truffle Blockchain using Ethereum 2.0, and a lightweight Proof of Authority (PoA) consensus algorithm demonstrates that the BACP-IeFC protocol significantly outperformed existing protocols in terms of average response time for edge device registration time, authentication time, and block preparation time.

In recent years, extensive studies have been carried out in community detection for social network analysis because it plays a crucial role in social network systems in today's world. However, most social networks in the real world have complex overlapping social structures, one of the NP-hard problems. This paper presents a new model for overlapping community detection that uses a multi-objective approach based on a hybrid optimization algorithm. In this model, the Modified Selection Function (MSF) hybrids the algorithms and recovery mechanism, the Slime Mould Algorithm (SMA), the Sine Cosine Algorithm (SCA), and the association strategy. Also, considering that these algorithms have been presented to solve single-objective optimization problems, the Pareto dominance technique has been used to solve multi-objective problems. In addition to overlapping community detection and increasing detection accuracy, the fuzzy clustering technique has been used to select the heads of clusters. Sixteen synthetic and real-world data sets were utilized to assess the suggested model, and the outcomes were contrasted with those of existing optimization techniques. The proposed model has performed better than the other tested algorithms in comparing the tests conducted by us in all 16 data sets, in the comparisons made with the algorithms proposed in other works in 11 data sets out of 14 data. The set has performed better than competitors. As a conclusion, the findings show that this model performs better than other methods.

In this paper, we introduce the large language model (LLM) ChatGPT-3.5 to automatically and intelligently generate a new metaheuristic algorithm (MA) according to the standard prompt engineering framework CRISPE (i.e., Capacity and Role, Insight, Statement, Personality, and Experiment). The novel animal-inspired MA named Zoological Search Optimization (ZSO) draws inspiration from the collective behaviors of animals for solving continuous optimization problems. Specifically, the basic ZSO algorithm involves two search operators: the prey-predator interaction operator and the social flocking operator to balance exploration and exploitation well. Furthermore, we designed four variants of the ZSO algorithm with slight human-interacted adjustment. In numerical experiments, we comprehensively investigate the performance of ZSO-derived algorithms on CEC2014 benchmark functions, CEC2022 benchmark functions, and six engineering optimization problems. 20 popular and state-of-the-art MAs are employed as competitors. The experimental results and statistical analysis confirm the efficiency and effectiveness of ZSO-derived algorithms. At the end of this paper, we explore the prospects for the development of the metaheuristics community under the LLM era.

Electric vehicles (EVs) have attracted significant attention in recent years primarily due to minimal adverse impact on the environment and efficiency of running costs. Although use of EVs brings noticeable benefits to users and the overall society, deployment of EVs, new carbon control regulations and interactive utility models are creating a compression on current electricity system. Further, due to the growth in adoption of EVs, the demand for electricity is expected to increase significantly over the next few years which can result in low-voltage networks. Since traditional networks are not designed for such loads, it can lead to inadmissible network conditions and resource overloads, which require network expansion through decentralized power generation. Distributed energy resources (DERs) such as smart grids leverage emerging technologies including internet of things (IoT) to achieve efficient energy management system. The aim of this research is to facilitate peer-to-peer energy distribution solution by focusing on the challenge of a decentralized, transparent reward system to achieve incentivization of power generation and distribution at microgrid level. Leveraging inherent benefits of blockchain technology, we develop a blockchain-based decentralized electricity trading platform to incentivize power generation at such micro level. Our platform allows local communities to contribute to meet the increased electricity demands by trading the generated electricity directly to EVs in a trusted and secure P2P environment while keeping the sustainability of the generated energy to balance demand and generation. We include detailed design specification, implementation and evaluation of the proposed electricity trading platform to assess feasibility of such system to be utilized within a production-level system.

IoT application uses MQTT, an application layer protocol that facilitates machine-to-machine communication using a central entity called broker. The vulnerability lies in the broker being susceptible to intrusion attempts, where a potential attacker might engage in a Distributed Denial of Service attack. Such an attack involves repetitively transmitting large number of malicious messages or counterfeit connect requests. To send large messages, the attackers must breach the authentication process of MQTT. MQTT employs two authentication approaches to safeguard its system: certificate-based and credential-based authentication. Credential-based authentication is popular as it is easy to implement. However, in MQTT, credential-based authentication is vulnerable to various attacks as credentials are transmitted in plain-text form. In literature, authors have explored different cryptography-based solutions to address these challenges. However, implementing these solutions in IoT systems is impractical due to the substantial computational requirements at the broker and the end devices. The primary objective of this work centres around formulating a PUF-based authentication policy and designing an IDS to track the behaviour of incoming traffic. In the proposed authentication scheme, the PUF mechanisms generate credentials to establish authenticity, thus protecting the network from password-based vulnerabilities like dictionary-based attacks. The second security module of this research implements a Machine Learning based IDS system to track and block fake connect requests in real-time. The proposed IDS system comprises Decision Tree and Neural Network algorithms that operate in parallel. In order to maintain the lightweight nature of the ML model, the system incorporates a feature selection technique. The result section shows that the proposed system effectively and efficiently recognizes fake connect requests in real-time and consumes minimal energy. Additionally, the proposed scheme requires less time than existing schemes in the literature.

Real-time and accurate traffic flow forecasting plays a crucial role in transportation systems and holds great significance for urban traffic planning, traffic management, traffic control, and more. The most difficult challenge is the extraction of temporal features and spatial correlations of nodes in traffic flow forecasting. Meanwhile, graph convolutional networks has shown good performance in extracting relational spatial dependencies in existing methods. However, it is difficult to accurately mine the hidden spatial-temporal features of the traffic network by using graph convolution alone. In this paper, we propose a multi-attention gated temporal graph convolution network (MATGCN) for accurately forecasting the traffic flow. Firstly, we propose a gated multi-modal temporal convolution(MTCN) to handle the long-term series of the raw traffic data. Then, we use an efficient channel attention module(ECA) to extract temporal features. For the complexity of the spatial structure of traffic roads, we develop multi-attention graph convolution module (MAGCN)including graph convolution and graph attention to further extract the spatial features of a road network. Finally, extensive experiments are carried out on several public traffic datasets, and the experimental results show that our proposed algorithm outperforms the existing methods.

Self-adaptive deep learning techniques provide scalability and flexibility in deploying and administrating deep learning models in the cloud environment. DL is widely used in cloud computing architecture, and these methods seek to optimize performance and resource utilization by automatically adjusting the resources allotted to machine learning tasks in response to workload fluctuations. Adaptive task scheduling algorithms maximise the distribution of DL techniques to available resources based on their features and needs. DL algorithms make intelligent judgements regarding job allocation, guaranteeing effective resource utilization and workload management. They consider variables, including task priority, resource availability, and resource capabilities. This research work deploys the Deep Convolutional Neural Network with a Fuzzy (DCNN-F) technique by differentiating the cloud nodes. The complexity of workflow scheduling in the cloud context is optimized by efficient learning, whereas energy and time consumption are effectively handled. The DCNN-F is trained with the resources in the cloud, and the solution for scheduling issues is rectified by learning data. The network is iteratively refined and optimized based on the feedback mechanism in DCNN-F. By combining the power of DCNN-Fs with efficient resource allocation strategies, research can maximise energy and time scheduling precedence-constrained tasks in cloud computing environments. The simulation outcome of DCNN-F is compared with state-of-art techniques, and DCNN-F outperforms Deep Q-Learning (DQL), Deep Reinforcement Learning based Optimization (DRL-O) and Deep Reinforcement Learning based Scheduling (DRL-S) techniques.