A Survey on Mobile Edge Computing: The Communication Perspective

IEEE Communications Surveys & Tutorials, Journal Year: 2017, Volume and Issue: 19(4), P. 2322 - 2358

Published: Jan. 1, 2017

Driven by the visions of Internet Things and 5G communications, recent years have seen a paradigm shift in mobile computing, from centralized cloud computing toward edge (MEC). The main feature MEC is to push network control storage edges (e.g., base stations access points) so as enable computation-intensive latency-critical applications at resource-limited devices. promises dramatic reduction latency energy consumption, tackling key challenges for materializing vision. promised gains motivated extensive efforts both academia industry on developing technology. A thrust research seamlessly merge two disciplines wireless communications resulting wide-range new designs ranging techniques computation offloading architectures. This paper provides comprehensive survey state-of-the-art with focus joint radio-and-computational resource management. We also discuss set issues, challenges, future directions research, including system deployment, cache-enabled MEC, mobility management green well privacy-aware MEC. Advancements these will facilitate transformation theory practice. Finally, we introduce standardization some typical application scenarios.

Language: Английский

---

Joint Service Caching and Task Offloading for Mobile Edge Computing in Dense Networks

IEEE INFOCOM 2022 - IEEE Conference on Computer Communications, Journal Year: 2018, Volume and Issue: unknown, P. 207 - 215

Published: April 1, 2018

Mobile Edge Computing (MEC) pushes computing functionalities away from the centralized cloud to network edge, thereby meeting latency requirements of many emerging mobile applications and saving backhaul bandwidth. Although existing works have studied computation of-floading policies, service caching is an equally, if not more important, design topic MEC, yet receives much less attention. Service refers application services their related databases/libraries in edge server (e.g. MEC-enabled BS), enabling corresponding tasks be executed. Because only a small number can cached resource-limited at same time, which cache has judiciously decided maximize performance. In this paper, we investigate extremely compelling but problem dynamic dense cellular networks. We propose efficient online algorithm, called OREO, jointly optimizes task offloading address key challenges MEC systems, including heterogeneity, unknown system dynamics, spatial demand coupling decentralized coordination. Our algorithm developed based on Lyapunov optimization Gibbs sampling, without requiring future information, achieves provable close-to-optimal Simulation results show that our effectively reduce for end users while keeping energy consumption low.

Language: Английский

---

Cloud Computing Resource Scheduling and a Survey of Its Evolutionary Approaches

ACM Computing Surveys, Journal Year: 2015, Volume and Issue: 47(4), P. 1 - 33

Published: July 21, 2015

A disruptive technology fundamentally transforming the way that computing services are delivered, cloud offers information and communication users a new dimension of convenience resources, as via Internet. Because provides finite pool virtualized on-demand optimally scheduling them has become an essential rewarding topic, where trend using Evolutionary Computation (EC) algorithms is emerging rapidly. Through analyzing architecture, this survey first presents taxonomy at two levels resources. It then paints landscape problem solutions. According to taxonomy, comprehensive state-of-the-art approaches presented systematically. Looking forward, challenges potential future research directions investigated invited, including real-time scheduling, adaptive dynamic large-scale multiobjective distributed parallel scheduling. At dawn Industry 4.0, for cyber-physical integration with presence big data also discussed. Research in area only its infancy, but rapid fusion technology, more exciting agenda-setting topics likely emerge on horizon.

Language: Английский

---

An Energy Efficient Ant Colony System for Virtual Machine Placement in Cloud Computing

IEEE Transactions on Evolutionary Computation, Journal Year: 2016, Volume and Issue: 22(1), P. 113 - 128

Published: Nov. 22, 2016

Virtual machine placement (VMP) and energy efficiency are significant topics in cloud computing research. In this paper, evolutionary is applied to VMP minimize the number of active physical servers, so as schedule underutilized servers save energy. Inspired by promising performance ant colony system (ACS) algorithm for combinatorial problems, an ACS-based approach developed achieve goal. Coupled with order exchange migration (OEM) local search techniques, resultant termed OEMACS. It effectively minimizes used assignment virtual machines (VMs) from a global optimization perspective through novel strategy pheromone deposition which guides artificial ants toward solutions that group candidate VMs together. The OEMACS variety problems differing VM sizes environments homogenous heterogeneous servers. results show generally outperforms conventional heuristic other evolutionary-based approaches, especially on bottleneck resource characteristics, offers savings more efficient use different resources.

Language: Английский

---

An overview of virtual machine placement schemes in cloud computing

Journal of Network and Computer Applications, Journal Year: 2016, Volume and Issue: 66, P. 106 - 127

Published: Jan. 29, 2016

Language: Английский

---

Cloud task scheduling based on ant colony optimization

Published: Nov. 1, 2013

Cloud computing is the development of distributed computing, parallel and grid or defined as commercial implementation these computer science concepts. One fundamental issues in this environment related to task scheduling. scheduling an NP-hard optimization problem, many meta-heuristic algorithms have been proposed solve it. A good scheduler should adapt its strategy changing types tasks. In paper a cloud policy based on ant colony algorithm compared with different FCFS round-robin, has presented. The main goal minimizing makespan given tasks set. Ant random search approach that will be used for allocating incoming jobs virtual machines. Algorithms simulated using Cloudsim toolkit package. Experimental results showed outperformed round-robin algorithms.

Language: Английский

---

Allocation of Virtual Machines in Cloud Data Centers—A Survey of Problem Models and Optimization Algorithms

ACM Computing Surveys, Journal Year: 2015, Volume and Issue: 48(1), P. 1 - 34

Published: Aug. 10, 2015

Data centers in public, private, and hybrid cloud settings make it possible to provision virtual machines (VMs) with unprecedented flexibility. However, purchasing, operating, maintaining the underlying physical resources incurs significant monetary costs environmental impact. Therefore, providers must optimize use of by a careful allocation VMs hosts, continuously balancing between conflicting requirements on performance operational costs. In recent years, several algorithms have been proposed for this important optimization problem. Unfortunately, approaches are hardly comparable because subtle differences used problem models. This article surveys formulations algorithms, highlighting their strengths limitations, pointing out areas that need further research.

Language: Английский

---

A Survey on the Placement of Virtual Resources and Virtual Network Functions

IEEE Communications Surveys & Tutorials, Journal Year: 2018, Volume and Issue: 21(2), P. 1409 - 1434

Published: Dec. 3, 2018

Cloud computing and network slicing are essential concepts of forthcoming 5G mobile systems. Network slices essentially chunks virtual connectivity resources, configured provisioned for particular services according to their characteristics requirements. The success cloud hinges on the efficient allocation resources (e.g., VCPU VMDISK) optimal placement virtualized functions (VNFs) composing slices. In this context, paper elaborates issues that may disrupt VNFs machines (VMs). This classifies existing solutions VM based nature, whether is dynamic or static, objectives, metrics. then proposes a classification VNF approaches, first, regarding general management VNFs, second, target type.

Language: Английский

---

Flower pollination algorithm: a comprehensive review

Artificial Intelligence Review, Journal Year: 2018, Volume and Issue: 52(4), P. 2533 - 2557

Published: March 13, 2018

Language: Английский

---

Efficient Task Offloading for IoT-Based Applications in Fog Computing Using Ant Colony Optimization

IEEE Access, Journal Year: 2020, Volume and Issue: 8, P. 37191 - 37201

Published: Jan. 1, 2020

The current thinking concerning computations required by Internet of Things (IoT) applications is shifting toward fog computing instead cloud computing, thereby achieving most the at network edge IoT devices. Fog can thus improve quality service delay-sensitive allowing such to take advantage low latency provided rather than high cloud. Therefore, tasks in various must be effectively distributed over nodes service, specifically task response time. In this paper, two nature-inspired meta-heuristic schedulers, namely ant colony optimization (ACO) and particle swarm (PSO), are used propose different scheduling algorithms load balance under communication cost time considerations. experimental results proposed compared with those round robin (RR) algorithm. evaluations show that ACO-based scheduler achieves an improvement times PSO-based RR balances nodes.

Language: Английский

---