The proliferation of the Internet of Things (IoT) and 5G technologies has established the edge-cloud continuum as a key architecture for high-performance, low-latency applications. However, effective task scheduling in these dynamic and heterogeneous environments presents significant challenges, including ensuring multi-hop network reliability under fluctuating conditions and optimizing queue management to prevent computational bottlenecks. To address these issues, this article proposes CPRGO, a novel task scheduling strategy based on an improved plant rhizome growth-based optimization algorithm enhanced with chaotic mapping. CPRGO enhances performance through three key mechanisms: it optimizes the selection of multi-hop transmission paths to boost task success rates, employs task classification and segmentation for more granular scheduling, and introduces a queue reorganization mechanism to mitigate performance degradation. Simulations performed on the RayCloudSim platform demonstrate that CPRGO significantly outperforms the baseline algorithms. It achieves superior performance in task completion rate, average cost, and average delay, demonstrating its robustness and practical applicability.

Rapid growth in mobile data traffic has led to an increased demand for efficient content delivery and data processing in wireless networks. In 5G networks, edge caching in Small-cell Base Stations (SBSs) is a promising technique that is capable of boosting the delivery of popular content by storing it closer to end users at the edge of the network. Adaptive edge caching refers to the process of dynamically selecting and caching content based on user demand and network conditions to further enhance the delivery of popular content. However, current solutions to Mobile Edge Computing (MEC) content caching are still limited and ineffective in many ways due to uncertainty in future content demands of mobile users, the distributed architecture of MEC networks, high operating costs and the contention among multiple users for shared network resources. In this paper, we propose a popularity-aware and physics-inspired approach for content caching in MEC by incorporating Newton's law of cooling for modeling the popularity of content, a self-feedback-enhanced proactive caching algorithm for periodically updating contents, and a multi-expert-based reactive caching algorithm for dynamically adapting to timevarying content request patterns. Experimental results based on real-world datasets demonstrate that our method outperforms benchmarks across multiple metrics.

In the conventional centralized networking architecture, content required by mobile users is obtained from distant Internet datacenters before being allocated on the mobile core network. This paradigm not only leads to content transmission delays but also imposes high bandwidth pressure on backhaul links, probably causing network congestion. Recently, cooperative edge caching technology is believed to be highly effective in addressing the above challenges. Nevertheless, it remains to be a great difficulty to properly decide how, when, and where to cache content over mobile edge computing (MEC) terminals. In this paper, we propose DCPS, a social Community-Interest-Aware Centralized Resource Scheduling Method. The proposed framework employs a Community Detection model and an Allocation (CDA) one to cluster users that are both socially connected and within geographic proximity. It assigns the communities formed by the clustering model to the most suitable base stations for content placement accordingly. Furthermore, it employs an Incremental Sequence Prediction Algorithm (IMSR) to predict user mobility as well as future content preferences for dynamic community adjustments with the help of a Reinforcement Learning-based Centralized Actor-Critic (CACP) algorithm. Simulations upon real-world datasets clearly demonstrate that the proposed algorithm outperforms existing methods across multiple performance metrics.

To meet the rapidly growing demands for services and applications in Mobile Edge Computing (MEC) environments, there is an increasing need to alleviate backhaul network pressure and enhance user experience. However, the dynamic nature of user mobility, fluctuating content popularity, and varying interest similarities within and across user communities pose significant challenges in designing efficient caching strategies. To address these challenges, this paper proposes a user interest-informed edge caching method by using a dynamic User-interest-based Clustring Caching (UCC) model. The proposed framework includes an improved density-based spatial clustering algorithm which employs the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm and a caching decision algorithm which adopts a federated deep reinforcement learning model for yielding high-quality and dynamic caching schedules. Experiments based on real-world urban taxi datasets and user interest datasets clearly demonstrates that the proposed method outperforms several existing algorithms across multiple performance metrics.

As vehicular networks shoulder an ever-expanding workload, smart task offloading has become essential for enhancing system throughput and maximizing resource efficiency. Nonetheless, the fluid topology, rapid node mobility, and varied user requirements make devising robust offloading schemes a persistent challenge. In this study, we proposes a task offloading method that synthesizes a reinforcement learning algorithm and an evolutionary one i.e., PPO-Enhanced NSGA-III Offloading Algorithm (PENOA). This method leverages a PPO reinforcement learning algorithm for dynamically adjusting offloading schedules produced by an NSGA-III algorithm. PENOA is capable of learning to generate near Pareto-optimal solution sets by collaboratively tuning, computational resource allocation and task schedules strategies between vehicles and servers. Experimental studies based on real-world Taxi Trajectory and Telecom Base Station datasets demonstrate that PENOA outperforms benchmark algorithms across multiple performance metrics.

In a Mobile Edge Computing (MEC) environment, as massive amounts of data are generated by user-end devices, the need to alleviate network backhaul pressure by offloading tasks to edge base stations closer to users is becoming increasingly critical. However, the dynamic nature of user mobility and varying interests within and across socially-connected user communities pose significant challenges in designing efficient pre-allocation strategies in MEC. In reality, MEC users can be socially connected and thus share common interests for task types. Consequently, we believe that the group interests of socially-connected MEC users can be exploited and propose a novel collaborative and group interest-informed resource pre-allocation, i.e., DeCoPre. It integrates the decentralized architecture that naturally divides the region into smaller zones for mitigating the impacts of the Single Point of Failure, a self-attention model for capturing multi-user interests and mobility patterns, a collaborative filtering approach for refining prediction results, and a grouping-based technique for resource pre-allocation algorithm. Numerical results upon real-world datasets clearly demonstrate that DeCoPre beats its peers across multiple performance metrics.

With the development of 5 G communication and Internet of Things (IoT) technology, increasing data is generated by a large number of IoT devices at edge networks. Therefore, increasing need for distributed Data Centers (DCs) are seen from enterprises and building elastic applications upon DCs deployed over decentralized edge infrastructures is becoming popular. Nevertheless, it remains a great difficulty to effectively schedule computational tasks to appropriate DCs at the edge end with low energy consumption and satisfactory user-perceived Quality of Service. It is especially true when DCs deployed over an edge environment, which can be highly inhomogeneous in terms of resource configurations and computing capabilities. To this end, we develop an edge task scheduling method by synthesizing a M/G/1/PR queuing model for characterizing the workload distribution and a Deep Deterministic Policy Gradient algorithm for yielding high-quality schedules with low energy cost. We conduct extensive numerical analysis as well and show that our proposed method outperforms state-of-the-art methods in terms of average task response time and energy consumption.

Mobile Edge Computing (MEC) is an emerging computing paradigm that offloads cloud center functions to the edge server. In a MEC environment, edge servers' limited storage and processing capacity require selective service caching, where only a part of required content can be placed directly upon the destination edge server and the remaining at remote cloud end. A primary challenge in this context is the creation of an effective and responsive service caching algorithm that improves the Quality of Service (QoS) perceived by users while reducing operational costs. This study applies an $M$ / G /1 queuing model as the foundational framework and transforms the service caching problem as an adversarial semi-bandit problem. We propose a delay-aware Genetic-Follow-the-Regularized-Leader (GFRL) algorithm, which is capable of guiding decentralized caching decisions. Experimental results indicate that GFRL outperforms traditional methods across various performance metrics.

Mobile edge computing(MEC) is an innovative technology that deploys computing resources around the demand side to provide near-request and responsiveness-guaranteed computing and storage services. A major attention paid by related works in this direction is mobility, where mobile traces of both edge users and servers are analyzed and exploited for accommodating offloading and migration requests for computation resources in a highly dynamic MEC environment. Our research in this work suggests that information of user interests, in terms of points of interest (POI), can be exploited in conjunction with mobility as well and proposes a hybrid method for for interest-informed and mobility-aware service migration path selection(HIMS). It synthesizes a trajectory prediction model and user interests prediction one for selecting target servers and reliable service migration paths. Experimental results demonstrate that our approach outperforms traditional methods across multiple performance metrics, especially those with sole input of mobility.

Mobile edge computing is an emerging computing paradigm boosting resource-demanding and delay-sensitive applications through deploying computing infrastructures at the edge of the Internet nearby mobile requesters and users. In an Internet of Vehicles (IoV) environment, Vehicular Edge Computing (VEC) is capable of exploiting network edge devices, in terms of, e.g., Roadside Units (RSUs), for predictive content caching for optimizing quality-of-experience (QoE) of nearby content requesters based on content popularity analysis, it remains a great challenge to accurately predict content popularity of mobile requesters and appropriately cache required content with low miss rate accordingly in a VEC environment with high user mobility and dynamics. To address this challenge mentioned above, in this paper, we propose predictive content popularity-aware approach, i.e., KM_SVD++, to edge caching in an VEC environment. The proposed approach is capable of achieving high hit rate of mobile content requestors in VEC and low latency of content delivery by leveraging a Kalman filtering model for predicting locations of vehicles and a SVD++ one for yielding decisions for cache deployment and replacement. We conduct extensive simulations as well to prove its effectiveness.

Mobile edge computing (MEC) provides edge services to users in a distributed and on-demand way. Due to the heterogeneity of edge applications, deploying latency and resource-intensive applications on resource-constrained devices is a key challenge for service providers. This is especially true when underlying edge infrastructures are fault and error-prone. In this paper, we propose a fault tolerance approach named DFGP, for enforcing mobile service fault-tolerance in MEC. It synthesizes a generative optimization network (GON) model for predicting resource failure and a deep deterministic policy gradient (DDPG) model for yielding preemptive migration decisions. We show through extensive simulation experiments that DFGP is more effective in fault detection and guaranteeing quality of service, in terms of fault detection accuracy, migration efficiency, task migration time, task scheduling time, and energy consumption than other existing methods.

Vehicle Edge Computing (VEC) is the deployment of applications close to edge servers to provide low latency and highly responsive services to users. However, due to the complexity and dynamics of the VEC environment, it is prone to errors and failures, and the reliability of edge service migration may be compromised if no measures are taken to cope with different levels of failures. To address this issue, this paper proposes an modified (m, n)-fault tolerance strategy (M-MNFT). Unlike the traditional one, which only considers ES failures, M-MNFT additionally selects redundant edge base stations to ensure task reliability during task migration, and takes into account the fact that the relative distance between the request and the base station is as small as possible when the request is sent, so as to avoid the impact of the edge base station failure on the Quality of Service (QoS) during task migration. In addition, we have performed extensive simulations to show that M-MNFT outperforms existing methods in terms of the number of delayed requests, on-time finish rate, and average waiting time.

The Quality-of-Service (QoS) aspects of Web service has gained popularity in the field of service computing. QoS-oriented Web service composition is a distributed model to construct new web service on top of existing primitive or other composite web services with QoS guarantees. A major challenge in this field is that the QoS data of candidate services are with run-time fluctuations and thus difficult to predict. Traditional approaches in this direction tended to address the challenge by statistics, prediction and neural network-based models. A major limitation of these methods lies in that they ignore outliers data in the historical QoS data, in terms of inconsistencies, errors, shifts, corruptions, etc. In this work, instead, we consider outliers in QoS series to be non-neglectable, and propose an outlier-tolerable and predictive approach to service composition through leveraging a joint estimation-based outlier detection method and a niched genetic algorithm. To validate the effectiveness of our proposed method, we conduct extensive case studies based on different outlier conditions, and the experimental results show that our method is superior to existing ones.

Edge computing is a decentralized computing infrastructure in which data, calculation, storage and applications are located somewhere between the data source and the computing facilities. While the edge servers enjoy the close proximity to the end-users to provide services at reduced latency and lower energy costs, we use from limitations in computational and radio resources, which calls for smart, quality-of-service (QoS) guaranteed and efficient task scheduling methods and strategies. For addressing the edge-environment-oriented multi-workflow scheduling problem, in this paper, we propose a probabilistic-QoS-aware approach to multi-workflow scheduling over edge servers with time-varying QoS. Our proposed method leveraged a probability-mass function-based QoS aggregation model and a discrete firefly algorithm for generating the multi-workflow scheduling plans. In order to prove the effectiveness of our proposed method, we conducted an experimental case study based on varying types of workflows and a real-world dataset for edge server positions. It can be seen that our method clearly outperforms its competitors in terms of completion time, cost, and deadline validation rate.

Most existing researches on cloud workflow systems have focused on resource scheduling with the aims to minimize system delay under budget constraints or optimize system cost under deadline constraints. However, cloud providers cannot guarantee a failure-free cloud environment, a compact scheduling plan is prone to failure, thus, workflow system reliability has been identified as a critical and challenging issue in the volatile cloud environment. With the ability of cloud, it is easy for users to implement the active fault tolerance schemes, e.g., Scale-Out. However, it will lead to issues like security problem and extra management cost. In this paper, we first investigate Scale-Up and Scale-Hybrid schemes to fully explore the possibilities offered by the ability of cloud. We formally model the problem of optimizing the reliability of a cloud workflow system under budget constraints with these three fault-tolerance schemes. These optimization problems are discrete and non-convex. Thus, we propose a genetic algorithm based method for workflow fault tolerance (GA4WFT). Finally, we evaluate the effectiveness and efficiency of proposed GA4WFT with three different fault-tolerance schemes through experiments conducted on Amazon EC2 data.

The cloud computing paradigm is characterized by the ability to provide flexible provisioning patterns for computing resources and on-demand common services. As a result, building business processes and workflow-based applications on cloud computing platforms is becoming increasingly popular. However, since real-world cloud services are often affected by real-time performance changes or fluctuations, it is difficult to guarantee the cost-effectiveness and quality-of-service (Qos) of cloud-based workflows at real time. In this work, we consider that workflows, in terms of Directed Acyclic Graphs (DAGs), to be supported by decentralized cloud infrastructures are with time-varying performance and aim at reducing the monetary cost of workflows with the completion-time constraint to be satisfied. We tackle the performance-fluctuation workflow scheduling problem by incorporating a stochastic-performance-distribution-based framework for estimation and optimization of workflow critical paths. The proposed method dynamically generates the workflow scheduling plan according to the accumulated stochastic distributions of tasks. In order to prove the effectiveness of our proposed method, we conducted a large number of experimental case studies on real third-party commercial clouds and showed that our method was significantly better than the existing method.

The cloud computing paradigm is featured by its ability to offer elastic computational resource provisioning patterns and deliver on-demand and versatile services. It's thus getting increasingly popular to build business process and workflow-based applications upon cloud computing platforms. However, it remains a difficulty to guarantee cost-effectiveness and quality of service of cloud-based workflows because real-world cloud services are usually subject to real-time performance variations or fluctuations. Existing researches mainly consider that cloud are with constant performance and formulate the scheduling decision-making as a static optimization problem. In this work, instead, we consider that scientific computing processes to be supported by decentralized cloud infrastructures are with fluctuating QoS and aim at managing the monetary cost of workflows with the completion-time constraint to be satisfied. We address the performance-trend-aware workflow scheduling problem by leveraging a timeseries-based prediction model and a Critical-Path-Duration-Estimation-based (CPDE for short) scheduling strategy. The proposed method is capable of exploiting real-time trends of performance changes of cloud infrastructures and generating dynamic workflow scheduling plans. To prove the effectiveness of our proposed method, we build a large-prime-number-generation workflow supported by real-world third-party commercial clouds and show that our method clearly beats existing approaches in terms of cost, workflow completion time, and Service-Level-Agreement (SLA) violation rate.

Server consolidation technique plays an important role in energy management and load-balancing of cloud computing systems. Dynamic virtual machine (VM) consolidation is a promising consolidation approach in this direction, which aims at using least active physical machines (PMs) through appropriately migrating VMs to reduce resource consumption. The resulting optimization problem is well-acknowledged to be NP-hard optimization problems. In this paper, we propose a novel merge-and-split-based coalitional game-theoretic approach for VM consolidation in heterogeneous clouds. The proposed approach first partitions PMs into different groups based on their load levels, then employs a coalitional-game-based VM consolidation algorithm (CGMS) in choosing members from such groups to form effective coalitions, performs VM migrations among the coalition members to maximize the payoff of every coalition, and close PMs with low energy-efficiency. Experimental results based on multiple cases clearly demonstrate that our proposed approach outperforms traditional ones in terms of energy-saving and level of load fairness.

Service composition is a technology capable of combing a collection of existing services where many smaller services are coordinated together to form a larger one. Functionally similar services can often show different quality-of-service (QoS) properties. For a specific service composition request, how to choose from a bag of suitable services that fulfill the required functions under given quality-of-service constraints is widely believed to be a great challenge. The traditional approach usually tackles this problem by assuming fixed, bounded, or statistic QoS and views the decision-making of service composition as a static process. Instead, we address this problem by considering time-varying and fluctuating QoS and presenting a predictive-trend-aware service composition method by using a time series prediction model and genetic algorithms. We conduct extensive case studies based on multiple randomly-generated service templates with varying process configurations and show that our method outperforms existing ones.

Virtual Machine (VM) consolidation technique plays an important role in energy management and load-balancing of cloud computing systems. Dynamic VM consolidation is a promising consolidation approach in this direction, which aims at using least active physical machines (PMs) through appropriately migrating VMs to reduce resource consumption. The resulting optimization problem is well-acknowledged to be NP-hard optimization problems. In this paper, we propose a novel merge-and-split-based coalitional game-theoretic approach for VM consolidation in heterogeneous clouds. The proposed approach first partitions PMs into different groups based on their workload levels, then employs a coalitional-game-based VM consolidation algorithm (CGMS) in choosing members from such groups to form effective coalitions, performs VM migrations among the coalition members to maximize the payoff of every coalition, and finally keeps PMs running in a high energy-efficiency state. The simulation results based on three scenarios clearly suggest that our proposed approach outperforms traditional ones in terms of energy-saving, and also achieve a fair level of load balance.