The Internet of Things (IoT) and its applications occupy a prominent place in contemporary research. The IoT, with its inherently heterogeneous behavior offers solutions to numerous problems and has become inseparable from human life. In this paper, an Optimized Multi-Channel Graph Neural Network Framework for Intrusion Detection in Internet-of-Things Environments (DGSEMGNN-IDF-IoT) is proposed. Here, the data collected from the Telemetry of Network-Internet of Things (ToN-IoT) dataset are used. To execute this, the input data are given into the pre-processing stage using the Dual Central Difference Kalman Filter (DCDKF) to eliminate redundant values and replace missing ones. The pre-processed data are supplied to the feature selection stage using the Ebola Optimization Search Algorithm (EOSA) to select the optimum features. The selected features are fed into the Dynamic Global Structure Enhanced Multi-Channel Graph Neural Network (DGSEMGNN) to effectively categorize the data as normal, backdoor, scanning, password, Cross-Site Scripting (XSS), ransomware, injection, Denial of Service (DOS), Distributed Denial of Service (DDOS) and Man-in-the-Middle (MITM) attack. Then, the Wolf-Bird Optimizer (WBO) is used to enhance the weight parameters of DGSEMGNN. The DGSEMGNN-IDF-IoT is implemented, and the performance metrics, such as accuracy, f1-score, specificity, sensitivity, Receiver Operating Characteristic (ROC), error rate, and computation time, are analyzed. The DGSEMGNN-IDF-IoT approach attains 6.80%, 10.41%, and 6.60% higher accuracy, 6.95%, 8.02%, and 10.9% higher f1-score when compared with existing methods.

Workload prediction is the necessary factor in the cloud data center for maintaining the elasticity and scalability of resources. However, the accuracy of workload prediction is very low, because of redundancy, noise, and low accuracy for workload prediction in cloud data center. In this manuscript, Workload Prediction in Cloud Data Centers using Complex-Valued Spatio-Temporal Graph Convolutional Neural Network Optimized with Gazelle Optimization Algorithm (CVSTGCN-WLP-CDC) is proposed. Initially, the input data is collected from two standard datasets such as NASA and Saskatchewan HTTP traces dataset. Then, preprocessing using Multi-Window Savitzky-Golay Filter (MWSGF) is used to remove noise and redundant the data. The preprocessed data is fed to CVSTGCN for workload prediction in a dynamic cloud environment. In this work, proposed Gazelle Optimization Approach (GOA) used to enhance the CVSTGCN weight and bias parameters. The proposed CVSTGCN-WLP-CDC technique is executed and efficacy based on workload prediction structure is evaluated using several performances metrics such as accuracy, recall, precision, energy consumption correlation coefficient, sum of elasticity index (SEI), root mean square error (RMSE), mean squared prediction error (MPE), and percentage prediction error (PER). The proposed CVSTGCN-WLP-CDC provides 23.32%, 28.53% and 24.65% higher accuracy; 22.34%, 25.62%, and 22.84% lower energy consumption when comparing to the existing methods using Artificial Intelligence augmented evolutionary approach espoused cloud data centres workload prediction architecture (TCNN-CDC-WLP), Performance analysis of machine learning centered workload prediction techniques for cloud (PA-BPNN-CWPC), Machine learning methods for effectual energy utilization in cloud data centers (ARNN-EU-CDC) methods respectively.

The expansion of the Internet of Things (IoT) ecosystem brings critical security risks, creating challenges for safeguarding enterprises and consumers. To enhance IoT security, this paper proposes an innovative cyber-attack detection framework that combines blockchain-based data preservation with deep learning. Data integrity is strengthened through a Proof of Transaction (PoTx)-based blockchain technique, securing data before classification. The Memory-Augmented Deep Unfolding Network (MADUN) is employed to identify diverse cyberattacks, including adversarial, poisoning, evasion attacks, service scanning, denial of service (DoS), keylogging, and data theft. Unlike conventional models, the MADUN classifier is optimized using the Bitterling Fish Optimization Algorithm (BFOA), achieving precise parameter tuning for improved classification accuracy. The experimental results show that the proposed MADUN-BC-CS-IoT framework outperforms existing approaches, such as DL-DCA-IoT, ICA-ML-IoT, and VEL-CAD-IoT in terms of accuracy, precision, recall, specificity, F1-score, RoC, and computational efficiency. These improvements underscore the efficacy of the MADUN-BC-CS-IoT technique in advancing secure IoT network operations.

Smart farming has entered a new phase made possible by the widespread deployment of cloud computing in the agricultural industry. Precision farming, improved crop management, and the promotion of sustainable agricultural practices are just a few of the ways in which cloud computing technologies are transforming the agricultural sector. Farmers are given an opportunity to make data-driven decisions and enhance resource usage, thanks to the scalability, accessibility, and cost-effectiveness of cloud-based solutions. In addition, cloud computing has several uses in smart farming, such as remote control of farming processes, real-time monitoring, predictive analytics, and data collecting from Internet of Things (IoT) devices. Cloud platforms have the ability to facilitate the collection and dissemination of data across ecosystem participants, hence promoting cooperation and the spread of knowledge in the agriculture sector. Concerns such as data security, privacy, and connection are discussed in this chapter as they pertain to implementing cloud computing in agriculture.

Internet of things (IoT) consists of wired/wireless network, sensor, and actuator, where security is more important when more devices are connected to IoT. To increase more security in IoT devices, this manuscript proposes a dual-channel capsule generation adversarial network (DCCGAN) espoused intrusion detection scheme for detecting security threats in IoT network (DCCGAN-IDF-DST-IoT). Data are collected from MQTT-IoT-IDS2020 dataset and Bot-IoT dataset. Then, the data are fed to local least squares, which eradicate the redundancy and replace the missing value. The pre-processed dataset is supplied to fertile field optimisation algorithm (FFOA), which selects the relevant features. Then DCCGAN is used for classifying the data as normal or anomalous. The proposed technique is activated in Python language. The performance of proposed technique for MQTT-IoT-IDS2020 dataset attains 16.55%, 21.37%, 32.99%, 27.66%, 26.45%, 21.47% and 22.86% higher accuracy compared with the existing methods.

Purpose Routing protocol for low-power lossy network (RPL) being the de facto routing protocol used by low power lossy networks needs to provide adequate routing service to mobile nodes (MNs) in the network. As RPL is designed to work under constraint power requirements, its route updating frequency is not sufficient for MNs in the network. The purpose of this study is to ensure that MNs enjoy seamless connection throughout the network with minimal handover delay. Design/methodology/approach This study proposes a load balancing mobility aware secure hybrid – RPL in which static node (SN) identifies route using metrics like expected transmission count, and path delay and parent selection are further refined by working on remaining energy for identifying the primary route and queue availability for secondary route maintenance. MNs identify route with the help of smart timers and by using received signal strength indicator sampling of parent and neighbor nodes. In this work, MNs are also secured against rank attack in RPL. Findings This model produces favorable result in terms of packet delivery ratio, delay, energy consumption and number of living nodes in the network when compared with different RPL protocols with mobility support. The proposed model reduces packet retransmission in the network by a large margin by providing load balancing to SNs and seamless connection to MNs. Originality/value In this work, a novel algorithm was developed to provide seamless handover for MNs in network. Suitable technique was developed to provide load balancing to SNs in network by maintaining appropriate secondary route.

On-demand computing ability and efficient service delivery are the major benefits of cloud systems. The limitation in resource availability in single data centers causes the extraction of additional resources from the cloud providers group. The federation scheme dynamically increases resource availability in response to service requests. The dynamic increase in resource count leads to excessive energy consumption, maximum cost, and carbon footprints emission. Hence, the reduction of resources is the major requirement to construct the optimized cloud source models for profit maximization without considering energy mix and CO2. This paper proposes the novel migration method to reduce carbon emissions and energy consumption. The initial stage in the proposed work is the categorization of data centers based on the MIPS and cost prior to job allocation offers scalable and efficient services and resources to the cloud user. Then, the job with the maximum size is allotted to the VM only if its capacity is less than the cumulative capacity of data centers. A novel migration based on overutilized and underutilized levels provides the services to the user even if the particular VM fails. The proposed work offers efficient maintenance of resource availability and maximizes the profit of the cloud providers associated with the federated cloud environment. The comparative analysis of the proposed algorithm with the existing methods regarding the response time, accuracy, profit, carbon emission, and energy consumption assures the effectiveness in a confederated cloud environment.

Workload prediction is the necessary factor in the cloud data center for maintaining the elasticity and scalability of resources. However, the accuracy of workload prediction is very low, because of redundancy, noise, and low accuracy for workload prediction in cloud data center. Therefore, in this article, a tree hierarchical deep convolutional neural network (T‐CNN) optimized with sheep flock optimization algorithm based work load prediction is proposed for sustainable cloud data centers. Initially, the historical data from the cloud data center is preprocessed using kernel correlation method. The proposed T‐CNN approach is used for workload prediction in dynamic cloud environment. The weight parameters of the T‐CNN model are optimized by sheep flock optimization algorithm. The proposed COSCO2 method has accurately predicts the upcoming workload and reduces extravagant power consumption at cloud data centers. The proposed approach is evaluated utilizing two benchmark datasets: (i) NASA, (ii) Saskatchewan HTTP traces. The simulation of this model is implemented in java tool and the parameters are calculated. From the simulation, the proposed method attains 20.64%, 32.95%, 12.05%, 32.65%, 26.54% high accuracy, and 27.4%, 26%, 23.7%, 34.7%, 36.5% lower energy consumption for validating NASA dataset, similarly 20.75%, 19.06%, 29.09%, 23.8%, 20.5% high accuracy, 20.84%, 18.03%, 28.64%, 30.72%, 33.74% lower energy consumption for validating Saskatchewan HTTP traces dataset than the existing approaches, like auto adaptive differential evolution algorithm BiPhase adaptive learning‐based neural network, error preventive score in time series forecasting models, time series forecasting methods for cloud data workload prediction, and self‐directed workload forecasting method.

The use of social distance as a tool throughout the battle with COVID-19 has shown promising results. In recent years, artificial intelligence (AI) and deep learning (DL) have emerged as a powerful resource for dealing with a wide range of practical challenges and delivering impressively positive outcomes. This article delves into the usage of Object Recognition, and Deep Learning to monitor personal and professional interactions between people at a distance. This research aims to aid in the pandemic fight by creating a technology that may be used as a social distance monitoring system. As evidence mounts in favour of social separation as a first-line, non-pharmaceutical intervention in the precautionary measure against the fastest spreading disease COVID-19, researchers feel that its rigorous observance should be encouraged. This proposal analyses live or recorded video to identify and evaluate correlation metric distances among individual people to verify to see if social distancing is preserved in crowded places.

The COVID-19 pandemic has been infecting the entire world over the past years. To prevent the spread of COVID-19, people have acclimatised to the new normal, which includes working from home, communicating online, and maintaining personal cleanliness. There are numerous tools required to prepare to compact transmissions in the future. One of these elements for protecting individuals from fatal virus transmission is the mask. Studies have indicated that wearing a mask may help to reduce the risk of viral transmission of all kinds. It causes many public places to take efforts to ensure that its guests wear adequate face masks and keep a safe distance from one another. Screening systems need to be installed at the doors of businesses, schools, government buildings, private offices, and/or other important areas. A variety of face detection models have been designed using various algorithms and techniques. Most of the articles in the previously published research have not worked on dimensionality reduction in conjunction with depth-wise separable neural networks. The necessity of determining the identities of people who do not cover their faces when they are in public is the driving factor for the development of this methodology. This research work proposes a deep learning technique to determine if a person is wearing mask or not and identifies whether it is properly worn or not. Stacked Auto Encoder (SAE) technique is implemented by stacking the following components: Principal Component Analysis (PCA) and Depth-wise Separable Convolutional Neural Network (DWSC-NN). PCA is used to reduce the irrelevant features in the images and resulted high true positive rate in the detection of mask. We achieved an accuracy score of 94.16% and an F1 score of 96.009% by the application of the method described in this research.

Cloud computing refers to a set of hardware and software that are connected to provide various computing services. Cloud consists of services to deliver software, infrastructure and platform over the internet depending on user demand. However, various categories of vulnerabilities and threats are increased, due to the improved demand and development of cloud computing. In cloud computing, data integrity and security are the main issues, which is reduced the system performance. Therefore, an Enhanced Elman Spike Neural Network (EESNN) based Fractional Order Discrete Tchebyshev moments (FrDTMs) encryption fostered big data analytical method is proposed in this manuscript for enhancing cloud data security. Initially, the input data is pre-processed using a Z-score normalization process. Also, the optimal features are implemented by Uni-variate ensemble feature selection technique, which can improve the quality and reliability of data. The proposed EESNN model is used to classify the attack types in cloud databases and the security is improved by FrDTMs model. The proposed EESNN-FrDTMs approach has achieved high data security and performance. The simulation of this model is done using JAVA. From the simulation, the proposed model attains 11.58%, 26.6%, 25.5%, 45.75%, 36.88% high accuracy, 7.37%, 15.43%, 8.68%, 11.42%, 16.88% lower information loss than the existing approaches, like PSO-PNN, eHIPF, GSNN-FEO, and BISNN-BO respectively.

The routing protocol for low power and lossy network (RPL) has tremendous scope in IoT, due to the fact that it can be customised as per network's domain requirement by altering objective function (OF). Many studies have proven that the careful crafting of OF with different metrics improve the quality of route identified by RPL. As the complexity of OF increases due to different metric combinations, the performance of RPL will start to deteriorate. We divide OF into a two-stage process to solve this problem. In the first stage we keep a simple OF that considers ETX and path delay and rank parents as per RPL specification. A fixed number of best ranked parents from the first stage OF are considered for further processing in the second stage OF. Fuzzy-based second stage OF considers mobility history, queue availability and remaining energy of the parents to select the preferred route. Second stage OF is run only on limited number of parents which improves performance of RPL. Performance evaluation shows that our two-stage objective function (TS-OF) reduces packet loss by 28% and energy consumption by 34% compared to the state of the art RPL objective function.

Increase in mobile nodes has brought new challenges to IoT’s routing protocol-RPL. Mobile nodes (MN) bring new possibilities as well as challenges to the network. MN creates frequent route disruption, energy loss and increases end-to-end delay in the network. This could be solved by improving RPL to react faster to route failures through route prediction, while keeping energy expenditure for this process in reasonable limits. In this context a new Mobility Energy and Queue Aware-RPL (MEQA-RPL) is proposed that have the capability to sense route failure and to identify proactively the next possible route before the current route fails. While identifying the next route, MEQA-RPL employs constraint check on energy and queue availability to guarantee QoS for MN and better lifetime for the network. When compared to RPL with mobility support our model reduce average signaling cost by 31%, handover delay by 32% and improve packet delivery ratio by 17%. We run simulations with multiple mobile nodes which have also shown promising results on aforementioned parameters.

Internet of Things will be inevitable in all walks of our life, where it becomes necessary for all smart devices to have end-to-end data transfer capability. These low power and low-cost end devices need to be enabled with IPv6 address and corresponding routing mechanism for participating in Internet of Things environment. To enable fast and efficient routing in Internet of Things network and constrained with limited energy, Routing Protocol for Low-power Lossy Network (RPL) has been developed by ROLL-Work Group. As RPL is proactive and energyconserving, it has become the most promising routing protocol for Internet of Things. Nodes in Low-power Lossy Network (LLN) are designed to conserve energy by maintaining radio silence over 90% of its lifetime. It is possible to further improve the node’s lifetime and thereby considerably extending network’s longevity by performing sensible routing. Different routing structure in Internet of Things network can be attained by carefully crafting the objective function for the same set of nodes which satisfies different goals. This paper focuses on different objective functions in RPL which have been developed over time with the emphasis on energy conservation and maximizing the lifetime of the network. In this work, we have carefully studied different metric compositions used for creating an objective function. The study revealed that combining metrics provides better results in terms of energy conservation when compared to single metric defined as part of RPL standard. It was also noted that considering some metric as a constraint can increase the rate of route convergence without affecting the performance of the network.

Nowadays, cloud computing makes it possible for users to use the computing resources like application, software, and hardware, etc., on pay as use model via the internet. One of the core and challenging issue in cloud computing is the task scheduling. Task scheduling problem is an NP-hard problem and is responsible for mapping the tasks to resources in a way to spread the load evenly. The appropriate mapping between resources and tasks reduces makespan and maximizes resource utilization. In this paper, we present and implement an independent task scheduling algorithm that assigns the users' tasks to multiple computing resources. The proposed algorithm is a hybrid algorithm for task scheduling in cloud computing based on a genetic algorithm (GA) and particle swarm optimization (PSO). The algorithm is implemented and simulated using CloudSim simulator. The simulation results show that our proposed algorithm outperforms the GA and PSO algorithms by decreasing the makespan and increasing the resource utilization. Indexing terms/

The concept of virtualization forms the heart of systems like the Cloud and Grid. Efficiency of systems that employ virtualization greatly depends on the efficiency of the technique used to allocate the virtual machines to suitable hosts. The literature contains many evolutionary approaches to solve the virtual machine allocation problem, a broad category of which employ Genetic Algorithm. This paper proposes a novel technique to allocate virtual machines using the Family Gene approach. Experimental analysis proves that the proposed approach reduces energy consumption and the rate of migrations, and hence offers much scope for future research.

Virtual machine (VM) allocation is the process of allocating virtual machines to suitable hosts. This problem is an NP-Hard problem. It can be considered as a variation of the bin-packing problem. Among various solutions that attempt to solve this problem, several approaches that apply Genetic Algorithm have been proposed. This paper proposes a method to improve the efficiency of such approaches. Implementation of the proposed approach shows significant improvements in the runtime, memory used, energy efficiency and SLA violations.