Efficient Computation Offloading of IoT-Based Workflows Using Discrete Teaching Learning-Based Optimization

Computers, materials & continua/Computers, materials & continua (Print), Год журнала: 2022, Номер 73(2), С. 3685 - 3703

Опубликована: Янв. 1, 2022

As the Internet of Things (IoT) and mobile devices have rapidly proliferated, their computationally intensive applications developed into complex, concurrent IoT-based workflows involving multiple interdependent tasks. By exploiting its low latency high bandwidth, edge computing (MEC) has emerged to achieve high-performance computation offloading these satisfy quality-of-service requirements devices. In this study, we propose an strategy for in a MEC environment. The proposed task-based consists optimization problem that includes task dependency, communication costs, workflow constraints, device energy consumption, heterogeneous characteristics addition, optimal placement tasks is optimized using discrete teaching learning-based (DTLBO) metaheuristic. Extensive experimental evaluations demonstrate effective at minimizing consumption reducing execution times compared strategies different metaheuristics, including particle swarm ant colony optimization.

Язык: Английский

---

A Computational Intelligence Hybrid Algorithm Based on Population Evolutionary and Neural Network Learning for the Crude Oil Spot Price Prediction

International Journal of Computational Intelligence Systems, Год журнала: 2022, Номер 15(1)

Опубликована: Авг. 21, 2022

Abstract This research attempts to reinforce the cultivating expression of radial basis function neural network (RBFnet) through computational intelligence (CI) and swarm (SI) learning methods. Consequently, artificial immune system (AIS) ant colony optimization (ACO) approaches are utilized cultivate RBFnet for approximation issue. The proposed hybridization AIS ACO (HIAO) algorithm combines complementarity exploitation exploration realize problem solving. It allows solution domain having advantages intensification diversification, which further avoids situation immature convergence. In addition, empirical achievements have confirmed that HIAO not only obtained best accurate theoretically standard nonlinear problems, it can be applied on instance solving practical crude oil spot price prediction.

Язык: Английский

---

A container deployment strategy for server clusters with different resource types

Concurrency and Computation Practice and Experience, Год журнала: 2023, Номер 35(10)

Опубликована: Март 7, 2023

Abstract The method of deploying microservices based on container technology is widely used in cloud environments. This can realize the rapid deployment and improve resource utilization datacenters. However, allocation container‐based are key issues. With continuous growth computing‐ storage‐intensive services, it necessary to consider different business types. study establishes a multi‐objective optimization problem model with similarity between containers servers, load balance clusters, reliability microservice execution as objectives. An improved artificial fish swarm algorithm proposed for microservices. comprehensive experimental results show that, compared existing strategies, matching degree server, cluster value, service reliability, other performance parameters while shortening running time algorithm. In addition, under constraint balancing, computing storage server clusters improved.

Язык: Английский

---

Joint Differential Evolution and Successive Convex Approximation in UAV-Enabled Mobile Edge Computing

IEEE Access, Год журнала: 2022, Номер 10, С. 57413 - 57426

Опубликована: Янв. 1, 2022

UAV-enabled mobile edge computing (MEC) is a emerging technology to support resource-intensive yet delay-sensitive applications with clouds (ECs) deployed in the proximity users and UAVs served as base stations air. The formulated optimization problems therein are highly nonconvex thus difficult solve. To tackle nonconvexity, successive convex approximation (SCA) technique has been widely used solve for by transforming objective functions constraints into suitable surrogates. However, optimal solutions based on approximated problem not original one they dependent feasible solution initialization. Unlike SCA, Differential Evolution (DE) global method that iteratively updates best candidate respect predefined functions. DE works well especially unconstrained since it can freely search very large regions of possible without considering convexity problem. when comes constrained problem, becomes inefficient find within given time limits. In view shortcomings incurred both we propose an innovative algorithm jointly applying SCA (DE-SCA) problems. directly using full initialize SCA-based will result worse function values often infeasible. Therefore, further design screen parts from utilize them algorithm. experimental simulations, consider system MEC where IoT devices, UAV ECs interact each other. simulation results demonstrate our proposed Screened DE-SCA largely outperforms benchmarks including DE, state-of-the-art algorithms system.

Язык: Английский

---

An Energy-Efficient Strategy and Secure VM Placement Algorithm in Cloud Computing

Computational Intelligence and Neuroscience, Год журнала: 2022, Номер 2022, С. 1 - 13

Опубликована: Авг. 25, 2022

One of the important and challenging tasks in cloud computing is to obtain usefulness by implementing several specifications for our needs, meet present growing demands, minimize energy consumption as much possible ensure proper utilization resources. An excellent mapping scheme has been derived which maps virtual machines (VMs) physical (PMs), also known machine (VM) placement, this needs be implemented. The tremendous diversity resources, tasks, virtualization processes causes consolidation method more complex, tedious, problematic. algorithm reducing use resource allocation proposed implementation article. This was developed with help a Cloud System Model, enables between VMs PMs among VMs. methodology used supports lowering number that are an active state optimizes total time taken process set (also makespan time). Using CloudSim Simulator tool, we evaluated assessed time. results compiled then compared graphically respect other existing energy-efficient VM placement algorithms.

Язык: Английский

---