Search

Containers have emerged as the most promising lightweight virtualization technology in providing cloud services due to its flexible deployment, portability, and scalability especially in micro-services, smart vehicles, IoTs, and fog/edge computing. An important and vital role in cloud container services is played by the scheduler's component to optimize performance and reduce cost due to the diverse nature of the workload and cloud resources. Despite the immense traction of containers in cloud computing, there is no comprehensive survey that covers container scheduling techniques. In this timely survey, we investigate the landscape of the state-of-the-art container scheduling techniques aiming to inspire more research work in this active area of research. The survey is structured around classifying the scheduling techniques into four categories based on the type of optimization algorithm employed to generate the schedule namely mathematical modeling, heuristics, meta-heuristics and machine learning. Then for each class of scheduling algorithms, we analyze and identify key benefits and pitfalls, together with key challenges of the available techniques based on the performance metrics. Finally, this paper highlights fertile future research opportunities to realize the full potential of the emergent container technology. [ABSTRACT FROM AUTHOR]

The cloud and the Internet of Things (IoT) provide optimal service responses at the user's request through edge devices. Resource sharing and computations are distributed between the users and edge devices, reducing the complex processing instances. In such computations, the pipelined process due to one-to-many resource sharing increases the service lags. For addressing this issue, this article presents a Service-Intensive Computation Offloading Framework (SICOF). The proposed framework guarantees service quality preventing lag in resource-oriented computations. This process is forwarded based on support vector classification for pipelined and distributed computations. The modifications in classifier intervals ensure distributed computations with offloading. In the offloading process, the service-free edge and cloud platform devices are identified, and the pipeline is split for validation. The classification outputs generate allocated request processing or service distribution, achieving a high response ratio. The service quality-based results are verified using the metrics service delay, computation time, process distribution ratio, and service lag. Simulation result shows that the proposed framework SICOF achieves reduce service delay, computation time, and service lag. [ABSTRACT FROM AUTHOR]

Sensor cloud provides a new way for the owners of different WSNs (Wireless Sensor Networks) to share their data to other parties (users/organizations). The edge-computing-assisted sensor-cloud systems (EC-assisted SCSs) can even supply real-time services for different parties. However, both the edge servers and the cloud servers in the systems are prone to be attacked in insecure network environments, since they are the bridges connecting both the end users and the WSNs in EC-assisted SCSs and attacking the servers can bring much more benefits for the attackers than just compromising some sensor nodes. In this paper, the problem of secure Top- k query in EC-assisted SCSs was studied, considering that non-trusted edge servers and non-trusted cloud servers exist in the systems, and a novel scheme was proposed to solve the problem. To achieve privacy preservation of the sensed data items and their scores, a secure key distribution protocol was designed, and some data-processing or integrity verification algorithms were also developed in the proposed scheme to preserve the data integrity of the query results. Thorough security analyses and extensive simulations show that the proposed scheme is efficient and secure, and it is perfectly fit for Top- k query in EC-assisted SCSs. [ABSTRACT FROM AUTHOR]

The traditional cloud computing paradigm faces challenges with the increasing number of Artificial Intelligence of Things (AIoT) devices generated at the network edge. Edge computing provides a novel approach to overcoming the limitations of cloud computing by delivering lower service latency and higher quality of service (QoS) to AIoT devices. However, edge computing encounters constraints due to scarce resources on edge servers and the long distance between AIoT devices and remote cloud servers. To ensure high QoS for AIoT devices, a balance is required between limited computing resources on the network edge and high latency caused by the geographic distance on the cloud side. In this paper, we propose an edge-cloud architecture that achieves optimal QoS for AIoT devices. We employ a federated learning-based architecture to train AIoT devices' data locally, thereby ensuring privacy. We evaluate the effectiveness and efficiency of our proposed approach by comparing it with the centralized approach from the state-of-the-art using two widely used datasets. The experimental results demonstrate that our architecture achieves higher effectiveness and efficiency in improving AIoT devices' QoS. Overall, our proposed edge-cloud architecture overcomes the limitations of traditional cloud computing, enhances user privacy, and delivers high QoS to AIoT devices. [ABSTRACT FROM AUTHOR]

The amount of data generated by internet of things devices multiplied day after day, and the use of cloud computing to store and process the data is very challenging due to the congestion of network, which negatively affects the processing time. The thing that led to the emergence of edge computing to solve these problems besides with the use of recent technologies such as artificial intelligence, blockchain and many more. The objective of this paper is to present the comparative study of the several proposed architectures on smart farming based on IoT according to different features since protocols, technologies, security, interoperability, accessibility, resilience etc., the advantages and limitations of each solution. Index Terms [ABSTRACT FROM AUTHOR]

The recent pandemic events in Thailand, Covid-19 in 2018, demonstrated the need for an event-based smart monitoring system. While a distributed multi-level architecture has emerged as an architecture of choice for a larger-scale smart event-based system that requires better latency, security, scalability, and reliability, a recently introduced data mesh paradigm can add a few additional benefits. The paradigm enables each district to become an event-based smart monitoring mesh and handle its analytics and monitoring workload. Districts can form a set of domains in a network of event-based smart community monitoring systems and provide data products for others during a crisis. This paper presents a distributed data mesh paradigm for an event-based smart monitoring product in a given community with predefined domains. The paper presents smart monitoring as a data product between domains. Key considerations for designing an event-based smart monitoring data product are given. The author introduces three possible domains necessary for creating a smart monitoring system in each community. Each domain creates a data product for a given domain and shares data between domains. Finally, a three-layer analytics architecture for a smart monitoring product in each domain and a use case is presented. [ABSTRACT FROM AUTHOR]

In dealing with emergency logistics, it is essential to use regional and central distribution hubs efficiently, and local and international sources of supply for automotive relief supplies are analyzed in this research. Resource scheduling algorithms employed by service providers to supply and assign resources in an environment are collectively referred to as resource scheduling. Cloud computing uses computational resources pooled and made available across a network, including information storage (cloud) and processing capabilities, requiring user participation in its deployment, operation, and maintenance. The challenging characteristics of such emergency logistics resource scheduling are inconsistencies in tracking, empty miles, and delivery delays. Hence, Emergency Supplier Distribution Mobile Edge Computing (ESD-MEC) research has been designed to improve emergency logistics resource scheduling algorithms in cloud computing. With the abovementioned requirements, ESD may address the distributed scheduling issue of vehicle emergency logistics resources. In particular, the MEC employs a specialized negotiating system to manage the scheduling of resources in impacted regions using an agent-based approach to ESD management in light of the need to do so. The proposed technology helps the decision-maker schedule resources in a dynamic environment and addresses supply demands. An ESD-MEC effectively predicts the rise in emergency logistics resources with faster vehicle strategies in cloud computing. The research concludes that ESD-MEC effectively indicates emergency logistics resource scheduling in cloud computing. The experimental analysis of vehicle logistics outperforms the method in terms of performance, accuracy, prediction ratio, mean square error, and efficiency ratio. [ABSTRACT FROM AUTHOR]

• Facilitating the Monitoring and Management of Structural Health in Civil Infrastructures. • An Edge/Fog/Cloud architecture for Structural Health Monitoring. • Unified interface to configure all the necessary steps in an SHM deployment. • Lightweight virtualization of components with containers for a portable distribution and scalability. • Performance analysis in a civil infrastructure with Fog and Cloud deployments. Structural Health Monitoring (SHM) is nowadays a requirement for civil infrastructures like tunnels, bridges and viaducts. With the advances provided by the Internet of Things (IoT) in recent years in areas such as wireless communications, miniaturisation and application protocols, this paradigm is spreading to the wider society to ensure appropriate protection of critical infrastructures. Both fog and edge computing have also strongly contributed to SHM and the IoT by bringing computing as close as possible to where data is produced, thereby reducing the latency response with respect to traditional cloud integrations. In this paper, an Edge/Fog/Cloud architecture for SHM in civil infrastructures is presented. The main goal of this architecture is to create and provide a flexible framework involving all the required components for the management, monitoring and deployment of SHM solutions, enabling high availability and easy distribution of the components over the architecture. The architecture has been evaluated as an alternative to the real deployment in a tunnel of a cloud architecture showing the benefits of adopting an Edge/Fog/Cloud hierarchy. [ABSTRACT FROM AUTHOR]

Vehicular networking allows vehicles with sensing capabilities to carry out communication from vehicle to vehicle or with accessible roadside units. To attain effective communication in vehicular networking, we have proposed a hybrid technological solution in terms of vehicular edge computing (VEC) utilizing resources, cloud and edge server environment. The proposed architecture supports a frequent communication between a vehicle and a legitimate edge server without any trusted party communication. Both security and privacy are two key challenges in edge based vehicular communications due to the adoption of a real-time environment that is required to respond to hundreds of moving vehicles. This paper presents an edge computing-based framework for vehicular communication to ensure security, message integrity, and privacy. The presented framework attains all security attributes in the presence of active as well as passive adversaries. The formal security proof ensures its correctness and robustness. The analysis indicates that it is secure and efficient, being based on new edge computing framework as compared to general cloud computing framework. [ABSTRACT FROM AUTHOR]

Edge computing has received wide attention due to the benefits it brings compared to the existing cloud computing model. With the emergence of this new paradigm, content service providers (CSPs) can extend their existing cloud-based CDN (content delivery network) to edge environments to achieve cost-effectiveness. Specifically, for those regions with many access requests, edge services can be rented for saving bandwidth costs. However, the cost of renting edge services is not negligible. If edge services are rented in places with few access requests, high renting costs will cause economic losses instead. Therefore, CSPs need to make rental decisions dynamically without any knowledge of the future. For solving this problem, we summarize it as an online edge service renting problem, and propose a randomized online edge-renting algorithm for CSPs to rent edge services cost-effectively. Through theoretical analysis, we prove that the cost of our algorithm would not exceed e / (e − 1 + α) times compared to the corresponding optimal algorithm, where α is the bandwidth discount of edge service compared to the cloud. Lastly, by verifying through experiments, the results show that our online algorithm can help CSPs save 11.9% of the total cost in edge service renting. [ABSTRACT FROM AUTHOR]

Electrocardiogram(ECG) monitoring is a critical and intricate task in cardiac healthcare. While large models supported by the remote cloud servers with abundant computational resources offers a feasible solution for this task, which still face challenges related to processing costs, privacy risks, and response time. Efforts have been made to integrate edge computing as a supplementary solution, but the current collaborative computing strategy for ECG monitoring is static and inefficient. To address these shortcomings, this study proposes an innovative confidence constraint cloud–edge collaborative computing(3CE2C) strategy. Firstly, the model implementation processes are illustrated, including an ultra-lightweight model for the edge node and a large model for the cloud server. To enhance classification performance, the personalized strategy is employed, resulting in an accuracy improvement from 0.9849 to 0.9929 in the model-cloud. For edge implementation, the optimal input length and model quantization both are explored to reduce the energy consumption. Based on the given confidence constraint, the models dynamically collaborate, with the low-confidence samples uploaded to the cloud server. This approach can achieve accuracy comparable to cloud computing, transmitting only about 17% low-confidence samples, the accuracy ratio(r A c c) is 0.9985. In addition, the method is validated in SVDB, where 3CE2C outperforms state-of-the art framework with the same uploaded sample ratio, resulting in a 1.54% improvement in classification accuracy. In conclusion, the proposed method provides a practical solution in real-time arrhythmia detection applications. • Proposing a confidence constraint cloud–edge collaborative computing strategy. • Improving the arrhythmia classification performance with a personalized strategy. • Implementing an ultra-lightweight model at edge with minimal performance loss. • Achieving accuracy similar to cloud computing, with uploading only 17% samples. • Demonstrating the superiority of the proposed framework compared to other studies. [ABSTRACT FROM AUTHOR]

The huge prospects of the internet of things (IoT) have led to an ever-growing demand for computing power by IoT users to enable various applications. Multi-access edge computing (MEC) research and development has rapidly gained attention during the last decade. The ability to deploy edge servers at different points across a content delivery network that can offer communication and computing services close to mobile user devices is one of the main factors driving the evolution of MEC. Furthermore, MEC has been considered a potentially transformational approach for fifth-generation (5 G) and beyond 5 G (B5G) networks, as well as a potential improvement to conventional cloud computing. Unmanned aerial vehicles (UAVs) can be used as effective aerial platforms to offer reliable and ubiquitous connections in wireless communication networks due to their distinctive qualities, such as high cruising altitude, on-demand deployment, and three-dimensional (3D) maneuverability. The number of research studies published in this area has dramatically increased due to the growing interest in UAV-enabled MEC. Although UAV-enabled MEC systems have been well studied, the existing models are becoming increasingly heterogeneous and scattered without harmony. This paper provides a comprehensive analysis of the literature on UAV-enabled MEC systems with a special focus on the system modeling, and optimization techniques for five identified domains, such as energy efficiency, resource allocation, trajectory control, latency, and security. For each domain, we have highlighted the recent advances, critical findings, and the advantages and disadvantages. Additionally, the identified proposed techniques were analyzed and discussed, with emphasize on the constraints and performance metrics. We also discuss a general system model for each highlighted domain. Moreover, the lessons are also derived from the study on system optimization and system modeling techniques identified in this paper. Then we discuss open issues related to UAV-enabled MEC systems in each highlighted domain, including problem formulation and optimization techniques. Finally, this paper lay out directions for future research to solve the aforementioned problems associated with UAV-enabled MEC systems. [ABSTRACT FROM AUTHOR]

The cloud–edge–client hierarchical computing architecture takes into account the dual advantages of cloud and edge computing. However, in realistic scenarios, the explosive increase of distributed data promotes the widespread application of the cloud–edge–client architecture, but also brings a variety of security threats. Relying on the advantages of federated learning technology, this has led to a growing interest in federated learning as the intersection of machine learning and cloud–edge–client fields. However, recent studies found that federated learning is potentially vulnerable to data poisoning attacks. Existing explorations of federated learning poisoning attacks and defense techniques in the cloud–edge–client environment are largely attack capabilities-specific and privacy protection-based : they are designed but fail to consider the characteristics of federated learning vulnerability and the difference in attack intensity, and lack flexibility. Yet few different attack capabilities and general defense strategies have been developed. To address these concerns, we introduce federated cloud–edge–client learning collaborative framework. Based on this architecture, a client-launched multi-label flipping attack method is discussed, which exhibits different attack capabilities. Further, we propose SPAD, a Server-based Parameter Adaptive Defense method in response to different attack intensities, to identify malicious attackers and benign participants, thereby improving the security of the federated learning image classification task. The results on image classification tasks demonstrate that in cloud–edge–client scenario, the multi-label flipping attack method presents different degrees of poisoning effect, and SPAD method has strong feasibility and effectiveness against poisoning attacks, which verify the significant advantages of the proposed method. • We explore the different attack intensities caused by MLFA methods under different settings, which are superior to non-MLFA. • Based on the FCECL collaborative framework, we explore the threats brought by the vulnerability of distributed learning. • We further propose a defense method, SPAD, which can adapt to poisoning attacks of different intensities. • Extensive experimental results demonstrate that MLFA and SPAD methods show outstanding advantages. [ABSTRACT FROM AUTHOR]

• IoT system modeling is necessary to find eligible deployments of offloading systems. • IoT systems modeling helps to drive the resource and performance optimizations. • IoT systems modeling alleviate limitations such as device resource constraints. • IoT and computing systems integration support emerging IoT application requirements. The continuous increase of IoT applications leads to a vast amount of data that needs to be transmitted, stored, and processed. Many IoT applications rely on the Cloud infrastructure to handle these specific application demands. However, the integration of IoT and Cloud poses challenges such as network delays, throughput, energy consumption, reliability, etc. Therefore, a new computing concept is required to support emerging IoT applications. These new concepts include fog computing, edge computing, mobile edge computing, mobile cloud computing, and cloudlets. They use various approaches to distribute resources, processes, and services among IoT system architecture layers. The challenge is to decide which offloading system is the best for a specific use case that emphasizes the IoT system modeling issue. In this paper, a model for the formal description of IoT systems is presented. In addition, an analytical evaluation method was proposed to design these systems using the corresponding architecture, technologies, protocols, and integration model to optimize performance. The proposed approach facilitates and simplifies the selection of the corresponding model for the system architecture. This approach enables an efficient method for performance optimization based on offloading processes (load balancing). Also, this paper provides some insights into specific emerging issues and ideas to be addressed by future research. [ABSTRACT FROM AUTHOR]

The increasing availability of Internet of Things (IoT) applications has led to the development of new technologies. Specifically, the deployment of edge servers close to IoT devices has strengthened the edge computing paradigm. With the collaboration of Mobile Edge Computing (MEC) and cloud computing, delay-sensitive and computation-intensive tasks can be offloaded to the edge/cloud servers to improve system performance in terms of the delay and energy consumption of IoT devices. However, there is a need to schedule the computation tasks for an efficient management. More importantly, the task scheduling strategy can face data tampering attacks to deliberately modify, destroy or manipulate the decisions. To solve the above problems, in this paper, we newly propose to integrate digital twin and blockchain into the edge networks. However, it is unclear (i) how energy and delay-aware computation should be combined, and (ii) which mining computations should be executed for a secure task scheduling. The state-of-the-art focuses on task scheduling and blockchain mining, separately. Therefore, we propose a novel blockchain-based digital twin-edge network architecture where our proposed algorithm solves these two challenges at the same execution. We design a three-layer system architecture, composed of physical entity layer, digital twin edge layer and blockchain layer. In the physical entity layer, we formulate an energy and delay-aware task scheduling problem. In the digital twin edge layer, we propose a novel Proof of Evaluation (PoE)-based secure energy and delay-aware task scheduling algorithm where optimization is executed by the genetic algorithm implementation of Warehouse Location Problem (WLP). In the blockchain layer, the best-found solutions are shared with the topology in a blockchain. Here, each block includes the hash of the previous block, a genetic algorithm-based solution, nonce value, and a hash of whole blocks in the blockchain. Thus, we aim to execute the computation tasks with an acceptable delay in an energy-efficient manner and prevent data tampering attacks against the optimal computation decisions. We validate the outcomes of our PoE-based secure digital twin-edge network model with extensive evaluations. Since the proposed model distributes the task not only to the local device but also to the MEC and cloud server for delay awareness, it reduces the delay but consumes more energy. Nevertheless, the additional energy consumption can be neglected against the delay reduction. The proposed scheme is also more scalable to compare with the conventional solution. The numerical results clearly show that the proposed model provides energy and delay awareness, maintaining both data integrity and trustworthiness at the same execution of algorithm. [ABSTRACT FROM AUTHOR]

A major objective of vehicular networking is to improve road safety and reduce traffic congestion. The experience of individual vehicles on traffic conditions and travel situations can be shared with other vehicles for improving their route planning and driving decisions. Nevertheless, the frequent occurrence of adversary vehicles in the network may affect the overall network performance and safety. These vehicles may behave intelligently to avoid detection. To effectively control and monitor such security threats, an efficient Trust Management system should be employed to identify the trustworthiness of individual vehicles and detect malicious drivers which is the major focus of this work. We propose a hybrid solution, which integrates Edge Computing and Multi-agent modeling in a Trust Management system for vehicular networks. The proposed solution also aims to overcome the limitations of the two commonly utilized approaches in this context: cloud computing and Peer-to-Peer (P2P) networking. Our framework has a set of features that make it an efficient platform to address the major security challenges in vehicular networks including latency, scalability, uncertainty, data accessibility, and malicious behavior detection. Performance of the approach is evaluated by simulating a realistic environment. Experimental results show that the proposed approach outperforms similar approaches from literature for various performance indicators. [ABSTRACT FROM AUTHOR]

To supervise massive generated data by the Internet of Things (IoT) efficiently, we face two issues that should be addressed which are: (1) heterogeneity or satisfying diversity among IoT devices, and (2) privacy-preserving or preventing unintentional disclosure of sensitive data. Through observation, we found that existing solutions apply one common privacy-preserving rule for all devices while they address the heterogeneity issue separately that lead to unappealing performance. In this paper, we propose a framework for addressing the heterogeneity issue and privacy-preserving of IoT devices at the network edge using a novel proposed ontology data model. Besides, it leverages the proposed ontology to obtain a privacy-preserving method by frequently changing the privacy-preserving behaviors of IoT devices. Through simulation, we show that our solution overhead is less than 9 percent in the worst situation so that it is affordable to most IoT devices in one of its applications that is smart city. [ABSTRACT FROM AUTHOR]

Deep neural networks are the state of the art for various machine learning problems dealing with large amounts of rich sensor data. It is often desirable to evaluate these models on edge devices instead of relying on cloud computing. In this paper, we perform traffic counting using surveillance cameras. Edge computing is required as only aggregated counts should leave the device and not the privacy sensitive video frames. Unfortunately, only small object detection models are suited for edge devices which results in sub-optimal performance. We introduce location specific models that are each trained for one specific camera. The model does not need to generalize to other locations. We show that smaller specialized models can outperform large general purpose models. We propose an automated way to train these small models without human intervention. We experimentally show that we can achieve a similar counting accuracy with 5x fewer parameters than state-of-the-art techniques. • We obtain high traffic counting accuracy in the real world using lightweight models. • We introduce location specific models that process the data of a single camera. • Since the models only need to work well for a single camera, they can be very small. • The location specific models can be trained automatically without human intervention. [ABSTRACT FROM AUTHOR]

The osmotic computing paradigm sets out the principles and algorithms for simplifying the deployment of Internet of Things (IoT) applications in integrated edge-cloud environments. Various existing simulation frameworks can be used to support integration of cloud and edge computing environments. However, none of these can directly support an osmotic computing environment due to the complexity of IoT applications and heterogeneity of integrated edge-cloud environments. Osmotic computing suggests the migration of workload to/from a cloud data center to edge devices, based on performance and security trigger events. We propose 'IoTSim-Osmosis– a simulation framework to support the testing and validation of osmotic computing applications. In particular, our detailed related work analysis demonstrates that IoTSim-Osmosis is the first simulation framework to enable unified modeling and simulation of complex IoT applications over heterogeneous edge-cloud environments. IoTSim-Osmosis is demonstrated using an electricity management and billing application case study, for benchmarking various run-time QoS parameters, such as IoT battery use, execution time, network transmission time and consumed energy. [ABSTRACT FROM AUTHOR]

The purposeful acts of bioterrorism and the emergence of new pathogens call for developing efficient biosurveillance systems that are capable of detecting health security threats and localizing their origins. For this sake, several biosurveillance systems have been developed and deployed in different regions around the world. The objectives of these systems are to efficiently detect threats and effectively manage responses. Although modern biosurveillance systems start to leverage recent advances in internet-of-things and communication technologies to detect natural biological outbreaks, it is not evident whether these systems can be generalized to detect threats related to purposeful biological attacks. As opposed to natural disease outbreaks, biological attacks can be triggered concurrently at different locations and in vital areas that are traversed by a large number of humans. In this paper, we propose an efficient framework to detect multiple concurrent threats and localize their origins based on real-time information and mobile-edge computing support. The proposed framework is organized into a hierarchical architecture. At the micro-level, monitored humans are equipped with wearable sensors that periodically capture the monitored human vital signs. To process the captured information efficiently, the monitored environment is partitioned into cells. Each cell is assigned a mobile-edge computing (MEC) component that is responsible for processing the monitoring information within its cell. The MEC components then send the processed information to a super component for final processing and to detect possible threats. The experimental results show that the proposed framework accurately detects multiple concurrent health security threats and localize their origins. [ABSTRACT FROM AUTHOR]

During the COVID-19 pandemic, the patient care delivery paradigm rapidly shifted to remote technological solutions. Rising rates of life expectancy of older people, and deaths due to chronic diseases (CDs) such as cancer, diabetes and respiratory disease pose many challenges to healthcare. While the feasibility of Remote Patient Monitoring (RPM) with a Smart Healthcare Monitoring (SHM) framework was somewhat questionable before the COVID-19 pandemic, it is now a proven commodity and is on its way to becoming ubiquitous. More health organizations are adopting RPM to enable CD management in the absence of individual monitoring. The current studies on SHM have reviewed the applications of IoT and/or Machine Learning (ML) in the domain, their architecture, security, privacy and other network related issues. However, no study has analyzed the AI and ubiquitous computing advances in SHM frameworks. The objective of this research is to identify and map key technical concepts in the SHM framework. In this context an interesting and meaningful classification of the research articles surveyed for this work is presented. The comprehensive and systematic review is based on the "Preferred Reporting Items for Systematic Review and Meta-Analysis" (PRISMA) approach. A total of 2540 papers were screened from leading research archives from 2016 to March 2021, and finally, 50 articles were selected for review. The major advantages, developments, distinctive architectural structure, components, technical challenges and possibilities in SHM are briefly discussed. A review of various recent cloud and fog computing based architectures, major ML implementation challenges, prospects and future trends is also presented. The survey primarily encourages the data driven predictive analytics aspects of healthcare and the development of ML models for health empowerment. [ABSTRACT FROM AUTHOR]

Due to technological advancement and deployment of handheld devices, mobile computing devices have gained the user attention to save resource usage, service access, and application use. In practice, it can be deployed widely to experience massive data storage and less computation cost. However, wireless channel access is vulnerable and insecure to prevent various potential attacks such as replay, redirection, etc. As a consequence, a seamless secure anonymous authentication scheme (S-SAAS) is proposed to create a secure session in cloud-based mobile edge computing. Since IoT, cloud computing, big-data and mobile technologies are converged to permit the group-based communication systems, it is highly demanded to strengthen the authentication process in order to reduce the security risk and vulnerabilities. Moreover, the experimental analysis proves that the proposed S-SAAS scheme consumes less signaling congestion and routing control overhead to achieve better service connection rate. [ABSTRACT FROM AUTHOR]

The tremendous growth in the internet of things (IoT) technology has provided great support for e-healthcare applications. The remote health monitoring of patients at home is gaining popularity with the increase in design of number of IoT enabled wearable devices. The secure and efficient health data retrieval during continuous uploading of huge data on network is a challenge. In addition, another challenge exists due to limited resources available with IoT device to proceed efficient data retrieval. On the other hand, the edge cloud in edge computing technology provide abundant resources to its devices for various computations, storage and processing. The integration of IoT with edge would solve the resource constraint issue. However, efficient retrieval of health data from large pool of data would still be an issue due to IP based solutions offered by IoT. The Named Data Networking (NDN) is a promising solution for efficient delivery of data based on its name based content searching. This paper proposed NDN communication based secure remote healthcare monitoring framework by integrating IoT with edge computing where security to the each individual content is provided through hashing and encryption mechanism. The performance of the proposed framework is evaluated against its companion in Icarus simulator. The results proved the efficiency of our framework for content retrieval delay and cost associated in accessing the required data. [ABSTRACT FROM AUTHOR]

Multi-access Edge Computing (MEC) offers cloud computing capabilities at the edge of the network. Growing demand for low-latency services requires Service Function Chains (SFCs) to be scaled up beyond MEC network to core network. To adapt to network dynamics and provide low-latency services, being able to migrate SFCs when needed is of paramount importance. However, migration of SFCs among edge and core networks such that average latency is optimized as well as considering resource consumption is an intractable challenge because improper migration of Virtual Network Functions (VNFs) results in failure of meeting the requirements of network policies. In this paper, we investigate SFCs in edge-core networks and model the Latency-aware Edge-Core SFCs Migration problem based on open Jackson networks. Two SFC migration algorithms, i.e., Profit-driven Heuristic Search (PHS) and Average Utilization Based (AUB), are proposed to efficiently optimize average latency of all SFCs in edge-core networks. Extensive evaluation results show that PHS optimizes average latency by 19.5%, while AUB can further reduce average latency by up to 36.9% by allowing a marginally higher number of VNF migrations. [ABSTRACT FROM AUTHOR]

In recent decades, artificial intelligence techniques have been adopted for many real-time applications. The Internet of Things (IoT) network comprises many sensing devices and physical objects for information gathering and further transmission. In addition to being sent to the receiving nodes, the collected data also needs to be received promptly. Also, many solutions have been proposed for IoT-based embedded systems using edge computing but they are not fully protected against unidentified communication threats. In such circumstances, such systems decrease the trust ratio, and communication performance is compromised. In this research, we describe an optimization model based on IoT-edged technology that incorporates cloud computational intelligence. Furthermore, edge nodes employ artificial intelligence algorithms to provide the optimal outcome for selecting trustworthy forwarded data and lengthen the connected time for smart devices. Firstly, the edge devices extract useful information from the IoT nodes, and accordingly, it provides a decision module based on optimization computing. Secondly, utilizing cryptographic approaches, edge technology secures the multi-layers of the IoT system and ensures data privacy with integrity. Finally, the proposed model is tested and verified for its performance than other related studies in terms of energy consumption, packet delivery ratio, and data delay. [ABSTRACT FROM AUTHOR]

In the world of cloud technologies, serverless computing has now settled as a stable and promising resident. This gives a cloud provider the flexibility to provide its users with both Platform-as-a-Service (PaaS), i.e., the back-end application runs in a dedicated container, or Function-as-a-Service (FaaS), i.e., the back-end logic is offered as elementary functions that are invoked by the client applications. In parallel, edge computing has attracted a significant interest, due its enticing promises of reducing the outbound traffic of telco operators, while at the same time cutting down the user latency. As a result, in the near future, PaaS and FaaS containers are going to cohabit in a versatile computation infrastructure spanning from the far edge up to the cloud. In this paper we propose a mathematical formulation of a resource allocation problem that optimizes the assignment of both types of containers and can be solved efficiently by an edge orchestrator. We evaluate the proposed solution via extensive simulation experiments, which show that our approach, which takes into account the characteristics of PaaS vs. FaaS, provides significant performance benefits compared to less sophisticated strategies, despite its relatively low run-time complexity. [ABSTRACT FROM AUTHOR]

Edge computing extends the Cloud computing paradigm by providing computation resources at the edge of the network where data is being generated to effectively meet the demands of real-time or latency-sensitive applications. Given Edge's particular characteristics, centralized controlling and decision making in the Cloud are no longer the only option. In contrast, it should be enabled throughout the Edge deployment's hierarchy from the top level of the hierarchy to near the IOT devices (at the edge of the network). This means enabling and enhancing intelligence and autonomy at the edge. This paper suggests an autonomous vision for Edge management. We propose a multi-agent system architecture, enabling autonomous decision making at Edge environments. A case study, using learning agents, is presented to illustrate the way the proposed solution enables sound management decisions. The case study focuses on smart offloading and autonomous power management. The profitability of enhancing intelligence and autonomy is proved in both cases. • Given Edge's particular characteristics, its control should be enabled throughout the Edge deployment's hierarchy from the top level of the hierarchy to near the IoT devices. • In Edge environments, the self-resource management is highly required since human intervention is impractical for such large Edge networks. • We argue not only for bringing computing resources closer to user devices and sensors as it was advocated by Edge computing, but also their management should be brought closer. • By calling up agent paradigm, we propose a multi-agent solution in order to deal with autonomous management at the Edge. • The profitability of enhancing intelligence and autonomy, by using learning agents, is proved in the case of smart offloading and autonomous power management. [ABSTRACT FROM AUTHOR]

The research aims to reduce the network resource pressure on cloud centers (CC) and edge nodes, to improve the service quality and to optimize the network performance. In addition, it studies and designs a kind of edge–cloud collaboration framework based on the Internet of Things (IoT). First, raspberry pi (RP) card working machines are utilized as the working nodes, and a kind of edge–cloud collaboration framework is designed for edge computing. The framework consists mainly of three layers, including edge RP (ERP), monitoring & scheduling RP (MSRP), and CC. Among the three layers, collaborative communication can be realized between RPs and between RPs and CCs. Second, a kind of edge–cloud​ matching algorithm is proposed in the time delay constraint scenario. The research results obtained by actual task assignments demonstrate that the task time delay in face recognition on edge–cloud collaboration mode is the least among the three working modes, including edge only, CC only, and edge–CC collaboration modes, reaching only 12 s. Compared with that of CC running alone, the identification results of the framework rates on edge–cloud collaboration and CC modes are both more fluent than those on edge mode only, and real-time object detection can be realized. The total energy consumption of the unloading execution by system users continuously decreases with the increase in the number of users. It is assumed that the number of pieces of equipment in systems is 150, and the energy-saving rate of systems is affected by the frequency of task generation. The frequency of task generation increases with the corresponding reduction in the energy-saving rate of systems. Based on object detection as an example, the system energy consumption is decreased from 18 W to 16 W after the assignment of algorithms. The included framework improves the resource utility rate and reduces system energy consumption. In addition, it provides theoretical and practical references for the implementation of the edge–cloud collaboration framework. [ABSTRACT FROM AUTHOR]

With the advent of the Internet of Everything, the combination of AI (Artificial Intelligence) and edge computing has become a new research hotspot, and edge intelligence has emerged, which enables network edge devices to analyze data through AI algorithms. Since the edge computing environment is more complex and variable than cloud computing, there are many issues in building edge intelligence, such as lack of quantitative evaluation criteria, heterogeneous computing platforms, complex network topologies, and changing user requirements. To analyze the performance of edge intelligence workloads running on heterogeneous hardware platform, we target machine learning workloads in edge intelligence and analyze the impact of algorithm model complexity, edge data characteristics and heterogeneous platform differences in edge intelligence in terms of relative performance. By analyzing the machine learning workload in edge intelligence, we find that the inference time and memory usage of a model can be predicted based on the amount of computation and number of parameters of the model. Moreover, image complexity, edge data network features, and batch size all affect the performance of edge intelligence workloads. Furthermore, the upper limit of model performance on the same computing platform is limited by hardware resources. And finally, the model performance of a platform depends on its own computing power and bandwidth. [ABSTRACT FROM AUTHOR]

In recent times, the Internet of Things (IoT) applications, including smart transportation, smart healthcare, smart grid, smart city, etc. generate a large volume of real-time data for decision making. In the past decades, real-time sensory data have been offloaded to centralized cloud servers for data analysis through a reliable communication channel. However, due to the long communication distance between end-users and centralized cloud servers, the chances of increasing network congestion, data loss, latency, and energy consumption are getting significantly higher. To address the challenges mentioned above, fog computing emerges in a distributed environment that extends the computation and storage facilities at the edge of the network. Compared to centralized cloud infrastructure, a distributed fog framework can support delay-sensitive IoT applications with minimum latency and energy consumption while analyzing the data using a set of resource-constraint fog/edge devices. Thus our survey covers the layered IoT architecture, evaluation metrics, and applications aspects of fog computing and its progress in the last four years. Furthermore, the layered architecture of the standard fog framework and different state-of-the-art techniques for utilizing computing resources of fog networks have been covered in this study. Moreover, we included an IoT use case scenario to demonstrate the fog data offloading and resource provisioning example in heterogeneous vehicular fog networks. Finally, we examine various challenges and potential solutions to establish interoperable communication and computation for next-generation IoT applications in fog networks. [ABSTRACT FROM AUTHOR]

Abstract Cloud based architectures present many challenges, ranging from poorly dimensioned hardware to lengthy wait times and resource balancing issues. Most of these problems can be mitigated by using proper performance evaluation techniques, depending on the accuracy of the model abstracting the real system. The adoption of a formalism over another to describe the considered infrastructure plays a crucial role in achieving this goal. In this sense, multiformalism proves itself to be a powerful modeling approach describing each system component according to the most suitable representation. This paper presents a novel modeling method oriented to the prediction of cloud architectures performance, suitable for joining the advantages of high level modeling abstractions and of the detail of a specialized simulator. Generalized Stochastic Petri Nets are used to describe the workload and the behavior of users and applications, while Cloudsim, a well known cloud infrastructure simulator, is adopted for the cloud part. A case study of a simplified Edge Computing application is presented to demonstrate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

With the rapid development of 5G network, business scenarios such as intelligent service and new retail are becoming more and more popular. The demand for more flexible and scalable real-time data processing, in particular, the AI-related data processing has also increased in edge computing. Therefore, how to meet such business development has become a major challenge. Focusing on this requirement, microservice architecture, proposed and developed by some big cloud computing companies' platform, such as Google Kubernetes platform, has gradually become a mainstream technology solution in edge computing. However, many microservices used in edge computing cannot achieve an even time distribution, which is random or sudden. Kubernetes built-in Horizontal POD Autoscaling (HPA) is unable to well handle the change of microservice load, which inevitably leads to the waste of system resources and affects the SLA of microservice. To solve this issue, this paper proposes a HANSEL system based on Kubernetes platform, which can optimize the horizontal elastic scaling policy of Kubernetes by accurately predicting the load of microservices based on the Bi-LSTM load prediction algorithm with attention mechanism. Furthermore, active elastic scaling is realized through reinforcement learning method, and we design a hybrid elastic scaling mechanism through combining reactive and active methods, so as to construct an elastic scaling system for automatic scheduling of working nodes. Our experimental results show that HANSEL system can improve the system resource utilization by about 20 % when meeting the microservice SLA of edge computing. [ABSTRACT FROM AUTHOR]

• Smart secured transportation is a pre-requisite for sustainable smart cities. • PKI-based authentication incurs high computation and communication overhead. • Successive authentication can reduce overhead in PKI-based authentication. • Edge computing concepts can enhance security and privacy in smart transportation system. Vehicular networks are getting huge attention among the research community due to its tremendous potential to offer diversified applications and services to the drivers and passengers towards the realization of intelligent transportation system (ITS) for connected living and sustainable smart cities. Cloud computing (CC), which possesses large capacity in terms of computing and storage resource, may not be suitable for many vehicular applications like safety and emergency services due to high bandwidth requirement and large response time. Edge or fog computing can be a promising technology that can be integrated with vehicular networks to meet the requirements of less bandwidth usage and lower response time. Vehicular cloud network (VCN) extends cloud computing concepts among the vehicles in the vehicular networks to enable vehicles with excess computing, storage or bandwidth capacities to share the resources with other vehicles, who are willing to use that extra resource to fulfill their service requirements. Security poses a great challenge in a vehicular network due to high mobility, dynamic and frequent changes in topology and sudden variation in vehicle density in the roads. Privacy of the vehicular users must be ensured to encourage mass participation for making the system trustworthy. Both security and privacy are pre-requisites of a sustainable smart city. For secured service provisioning, before transmitting a message, connected vehicles must be authenticated. In this paper, we have proposed a system to authenticate vehicles for secured service provisioning in edge-enabled vehicular cloud networks. A novel technique has been developed to avoid redundant authentication to validate a vehicle successively across various RSUs (road side units). Extensive simulation works have been done and the outputs demonstrate efficacy of the proposed algorithm. [ABSTRACT FROM AUTHOR]

• Cloudlet is an important part of providing cloud services in Mobile Edge Computing (MEC) with sustainability. • The cloudlet is providing cloud services with low latency. Hence the task allocation and scheduling inside of the cloudlet is a challenging job. • Heuristic load balancing strategy, which minimizes task completion time and makespan enhances efficiency of cloud services. • The dynamic task allocation strategy is proposed, this approach is compared with the queue-based task allocation strategy. • The proposed method outperforms the standard algorithm in most of the observed cases with sustainability. Cloudlet is an important part of providing cloud services in Mobile Edge Computing (MEC) with sustainability. As the number of mobile users grows rapidly in the current era, the load in the cloudlet becomes very high. The cloudlet is considered in the middle layer for providing cloud services with low latency and energy efficiency. Hence the task allocation and scheduling inside of the cloudlet is a challenging job. In the recent past, many research works conducted without considering real-time network parameters. In this work, a heuristic load balancing strategy is designed and analyzed to minimize the task completion time and makespan, which enhances the efficiency of cloud services. The proposed method is considered a dynamic task allocation strategy to the cloudlet concerning network bandwidth, network delay, energy efficiency, and this approach is compared with the queue-based task allocation strategy. The experiment is conducted to compare the proposed method with the recently developed standard algorithms over the synthetic dataset. Experimental results of the proposed work show a significant improvement in task scheduling in terms of energy and execution cost, as well as time, compared to the existing methods. The proposed method outperforms the standard algorithm in most of the observed cases with sustainability. [ABSTRACT FROM AUTHOR]

This paper provides a detailed description of a new simulator, PFogSim, developed to facilitate testing of fog computing configurations of various sizes with fog nodes distributed over large geographical areas and interconnected over one or more network links. PFogSim allows for a multi-layered fog environment design as opposed to existing single or two-layered designs. Furthermore, it allows fog nodes and network links of different capacities, thus enabling realistic, network configurations to be tested. PFogSim fills the void in which simulators for very large, practical fog networks are lacking. © 2021 Elsevier Science. All rights reserved. • Proposed and implemented the PFogSim simulator for large-scale heterogeneous fog computing environments. • Simulates large-scale multilayered hierarchical fog environments having continuous mobility of devices, users, fog nodes. • Simulates cloud/edge/mobile computing deployments by restricting the number of fog layers and node resource configurations. • Newer version has features such as, 3D locations for devices/users/nodes, their energy consumption modeling, etc. [ABSTRACT FROM AUTHOR]

Mobile Edge Computing (MEC) enables smart mobile devices to handle computationally intensive tasks with limited resources. By offloading tasks to the edge of the network closer to mobile devices, mobile devices can provide users with intelligent services with shorter latency and lower energy consumption than cloud computing. while mobile edge computing is superior in terms of reliability and low latency, it is still not a complete replacement for cloud computing. In order to make full use of the large amount of network resources of cloud computing as well as to take advantage of parallel computing, this paper investigates the partial offloading of multiple mobile devices in the cloud-side collaboration scenario. Through ​​game theory, we first obtain the edge-to-terminal optimal offloading decision, then find the cloud-edge optimal offloading decisions for all terminal devices that choose to offload to the edge, and finally reach the goal of optimal offloading on the three sides of the cloud-edge-terminal. Based on this idea, we propose a two-stage computing offloading algorithm (TSGTA). The whole algorithm is divided into two stages, and both stages are designed to offload algorithms based on game theory. The simulation results show that under the time-delay constraint, this method has an average improvement of 32% compared with the traditional method. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT), including wireless sensors, is one of the highly anticipated contributors to big data; therefore, avoiding misleading or forged data gathering in cases of sensitive and critical data through secure communication is vital. However, due to the relatively long distance between remote cloud and end nodes, cloud computing cannot provide effective and direct management for end nodes, which leads to security vulnerabilities. In this paper, we propose a novel trust evaluation model based on the trust transitivity on a chain assisted by mobile edge nodes, which is used to ensure the reliability of nodes in the Internet of Things and prevent malicious attacks. The mobile edge nodes offer a new solution to solve the above problems with relatively strong computing and storage abilities. Firstly, we design calculation approaches to different trust chains to measure their trust degrees. Secondly, we propose an improved Dijkstra's algorithm for collecting trust information of sensor nodes by mobile edge nodes. Finally, the experimental results show that our trust model based on mobile edge nodes can evaluate sensor nodes more precisely and enhance the security on the Internet of Things. • An efficient trust chain evaluation model assisted by mobile edge nodes is proposed. • The model provides calculation methods of the value of the different trust chains. • Trust information of sensor nodes are collected by an improved Dijkstra algorithm. [ABSTRACT FROM AUTHOR]

A co-simulation may comprise several heterogeneous federates with diverse spatial and temporal execution characteristics. In an iterative time-stepped simulation, a federation exhibits the Bulk Synchronous Parallel (BSP) computation paradigm in which all federates perform local operations and synchronize with their peers before proceeding to the next round of computation. In this context, the lowest performing (i.e., slowest) federate dictates the progression of the federation logical time. One challenge in co-simulation is performance profiling for individual federates and entire federations. The computational resource assignment to the federates can have a large impact on federation performance. Furthermore, a federation may comprise federates located on different physical machines as is the case for cloud and edge computing environments. As such, distributed profiling and resource assignment to the federation is a major challenge for operationalizing the co-simulation execution at scale. This paper presents the Execution Performance Profiling and Optimization (EXPPO) methodology, which addresses these challenges by using execution performance profiling at each simulation execution step and for every federate in a federation. EXPPO uses profiling to learn performance models for each federate, and uses these models in its federation resource recommendation tool to solve an optimization problem that improves the execution performance of the co-simulation. Using an experimental testbed, the efficacy of EXPPO is validated to show the benefits of performance profiling and resource assignment in improving the execution runtimes of co-simulations while also minimizing the execution cost. [ABSTRACT FROM AUTHOR]

The user demand increment and resource restrictions in IoT can provide efficient services without considering the service cost and service quality. Nowadays various computing technologies like fog, cloud, and edge computing and their confluence of IoT are required. The standard IoT services can be transformed into composite services via the incorporation of cloud computing with IoT. These IoT composite services are not suitable for various QoS scenarios which results in performance degradation and exhibited real-time adaptive sensing issues. Therefore, a novel Deep Q-learning Network-based Hybrid Enhanced Chimp Artificial Hummingbird (DQN-Hybrid ECAH) approach is proposed to overcome the state-of-art approach problems. The DQN-Hybrid ECAH approach represents a significant novelty in the field of IoT, as it utilizes a combination of advanced techniques to overcome the challenges posed by composite services in IoT. In particular, the use of deep Q-learning networks allows for the optimization of composite service quality, while the enhanced chimp artificial hummingbird techniques ensure that the approach is both efficient and effective. From the experimental evaluation, it is evident that the proposed approach is superior to other state-of-the-art methods. Specifically, the packet delivery ratio (PDF) and throughput rate obtained were high, with minimum jitter and average end-to-end delay. This demonstrates the effectiveness of the DQN-Hybrid ECAH approach in enhancing composite service quality in IoT applications. Overall, the proposed DQN-Hybrid ECAH approach represents a significant contribution to the field of IoT, providing a novel solution to the challenges associated with composite services in IoT. The approach has the potential to significantly improve the performance of IoT applications, while also reducing the real-time adaptive sensing issues that are often encountered in this context. [ABSTRACT FROM AUTHOR]

Thermal error compensation is one of the most research-oriented topics in manufacturing with rising importance in the industry. This paper presents an innovative Industry 4.0 application of thermal error compensation for precision engineering. A federated learning-based thermal error compensation approach running in the cloud is applied to two machine tools, one located at ETH Zürich, and another one at TU Wien. Although environmental conditions and thermal error behaviour of both machines differ, the implemented knowledge transfer across machines is a viable compensation strategy, albeit with limited precision. A detailed comparison of the two machines of the same type under the same load conditions shows foreseeable similarities in behaviour, but also clear differences due to the different configurations and lifetime status. The cloud-based compensation reduced the crucial thermal errors in the best case of both machine tools by more than 80% under critical conditions. • A new cloud-based thermal error compensation approach is introduced. • Federated learning in terms of cloud and edge computing for reaching a high level of safety and security is presented. • Detailed analysis of the same type of machine tool with different wear and slightly different configurations. • The transferability of the model calibration period (learning phase) is reached by model transfer (transfer learning). • Compensation results already reduce the thermal error by up to 80 %. [ABSTRACT FROM AUTHOR]

Thanks to the explosive evolution of 6G, Artificial intelligence, Blockchain, Cloud computing, Data, and Edge computing (6G+ABCDE), it is possible to make Intelligent Transportation Systems (ITS) more autonomous, intelligent, and efficient. Conventional ITS technologies have been applied and deployed worldwide, and autonomous vehicles have also made a qualitative leap. The revolution of ITS magnitude is on the horizon, and the singularity may happen in the 6G Autonomous Intelligent Transportation Systems (6G-AITS). We indicate that the human-centric transportation system will still be the key to 6G-AITS. The 6G-AITS is a three-dimensional intelligent system, which is driven by data and intelligence, and characterized by sharing and collaboration, thereby achieving the service-oriented demand response. To support this vision, we explore the 6G-AITS changes in terms of system architecture, communication, and resource allocation, as well as its key features. We elaborate on the mechanism and three applications of 6G-AITS. Furthermore, we explore the challenges of 6G-AITS and consider some issues that go beyond technologies that could affect the development of 6G-AITS. Key objective of this work is to provide an exploration of the imperative requirements to develop 6G-AITS and guidance for the research and development of 6G-AITS. [ABSTRACT FROM AUTHOR]

Internet of Things (IoT) devices have gained more prevalence in ambient assisted living (AAL) systems. Reliability of AAL systems is critical especially in assuring the safety and well-being of elderly people. Runtime verification (RV) is described as checking whether the observed behavior of a system conforms to its expected behavior. RV techniques generally involve heavy formal methods; thus, it is poorly utilized in the industry. Therefore, we propose a democratization of RV for IoT systems by presenting a model-based testing (MBT) approach. To enable modeling expected behaviors of an IoT system, we first describe an extension to a UML profile. Then, we capture the expected behavior of an interaction that is modeled on a Sequence Diagram (SD). Later, the expected behaviors are translated into runtime monitor statements expressed in Event-Processing Language (EPL), which are executed at the edge of the IoT network. We further demonstrate our contributions on a sample AAL system. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) has become the forefront of bridging different technologies together. It brings rise to online computational services that make mundane tasks convenient. However, the volume of devices connecting to the network started to increase. In turn, services that thrived on centralized storage are being strained and overloaded. As applications and software advances, processing and computational power become a concern to technology companies. With data risks and large numbers of connected devices, cloud computing has become outdated. Devices are forced to commit unnecessary expenses to stay relevant in the market due to the increase in software complexity. This need for change resulted in the introduction of edge computing. Edge computing distributes the computational strain between the server and the devices. This contribution allows the cloud to accommodate more users and devices are no longer in need to make significant changes to their design every so often. Many real-time applications have evolved to require high amounts of processing power to execute. For example, sound classification comes with massive computational needs due to its affiliation with neural networks and deep learning. This paper aims to create a feasible and deployable real-time sound classification system. There were three configurations tested in this paper. The results of our experiments show that cloud computing and edge computing alone cannot cater to a technological market that is exponentially growing in size and complexity. However, the same results show promise in finding optimal configurations in terms of a combination of end device power consumption, application runtime and server latency to systems instead of focusing on a single model. Overall, it is better to take into consideration the strengths and weaknesses of each computing architecture. In finding a reasonable configuration balance, lower power consumption in end devices and lesser computational strain on cloud servers is a must. [ABSTRACT FROM AUTHOR]

The rapid development of cloud computing and the Internet of Things (IoT) have facilitated near real-time optimization of the group distributed manufacturing systems. Currently, the most common technique to accomplish near-real-time optimization is cloud–edge cooperation for offloading optimization tasks. The tasks are partially offloaded to the cloud to be completed, and the remaining are kept at the edge. Due to the complexity of task offloading, such as capacity restrictions of cloud and edge computing resources, or task deadlines, unbalanced or insufficient tasks are offloaded to cloud and edge, causing time delay. To address the imbalance and insufficiency in the task offloading process, a mixed-integer programming model was developed to reduce the latency of task calculation. The task offloading problem is decomposed into two sub-problems: 1) Defining priorities for the tasks in near real-time. 2) Determining if the task is offloaded to the cloud. A multi-agent deep reinforcement learning with attention mechanism (MaDRLAM) framework is proposed to solve the two-step decision problem. The MaDRLAM framework consists of two agents, and each agent corresponds to a sub-problem. Each agent comprises an encoder and a decoder, and the two agents cooperate in devising an offloading strategy for the tasks. The Encoder and Decoder built for each agent are based on the Transformer structure. Unlike the traditional Transformer, we added the Pointer networks to the Transformer to solve the proposed decision problem. Besides, an improved multi-actor and single-critic strategy based on the REINFORCE algorithm is designed to train the proposed MaDRLAM. Finally, Extensive computational experiments are conducted on instances with a varying number of tasks, different task data sizes, and different cloud computing capacities. Computational results show that the proposed framework can find a solution with a GAP value of less than 1% within 1 s for each instance. The proposed framework is competitive in both solution accuracy and solution time compared with other offloading strategies. [ABSTRACT FROM AUTHOR]