Workload and Capacity Optimization for Cloud-Edge Computing Systems with Vertical and Horizontal Offloading

Thai, Minh-Tuan; Lin, Ying-Dar; Lai, Yuan‐Cheng; Chien, Hsu-Tung

doi:10.1109/tnsm.2019.2937342

Cited by 60 publications

(20 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In problem solving, some scholars use traditional methods. Thai et al [6] proposed a generic architecture for cloud edge computing with both vertical and horizontal service computing nodes, formulated as a mixed-integer nonlinear programming problem, proposed an approximation algorithm to iteratively obtain the optimal solution using a branch-and-bound method to minimize the system computation and communication costs. Cui et al [7] made a trade-off between energy consumption and latency to meet the user requirements of different IoT applications, formalized the problem as a constrained multiobjective optimization problem, and found the optimal solution by an improved fast elite non-dominated ranking genetic algorithm.…”

Section: Offloading Methods With Different Problem Solving Strategiesmentioning

confidence: 99%

“…At present, researchers have conducted a lot of research on the offloading decision problem, such as seeking the optimal solution after formulating the problem based on traditional algorithms [6][7][8][9][10], maximizing the reward to make appropriate decision actions based on reinforcement learning algorithms [11][12][13][14], etc. However, the above algorithms do not take the dynamic and predictable nature of the task during task offloading into account, thus they may produce unreasonable offloading decisions, resulting in degradation of the task offloading performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Task Offloading Based on LSTM Prediction and Deep Reinforcement Learning for Efficient Edge Computing in IoT

Chen

Yan

et al. 2022

Future Internet

View full text Add to dashboard Cite

In IoT (Internet of Things) edge computing, task offloading can lead to additional transmission delays and transmission energy consumption. To reduce the cost of resources required for task offloading and improve the utilization of server resources, in this paper, we model the task offloading problem as a joint decision making problem for cost minimization, which integrates the processing latency, processing energy consumption, and the task throw rate of latency-sensitive tasks. The Online Predictive Offloading (OPO) algorithm based on Deep Reinforcement Learning (DRL) and Long Short-Term Memory (LSTM) networks is proposed to solve the above task offloading decision problem. In the training phase of the model, this algorithm predicts the load of the edge server in real-time with the LSTM algorithm, which effectively improves the convergence accuracy and convergence speed of the DRL algorithm in the offloading process. In the testing phase, the LSTM network is used to predict the characteristics of the next task, and then the computational resources are allocated for the task in advance by the DRL decision model, thus further reducing the response delay of the task and enhancing the offloading performance of the system. The experimental evaluation shows that this algorithm can effectively reduce the average latency by 6.25%, the offloading cost by 25.6%, and the task throw rate by 31.7%.

show abstract

Section: Offloading Methods With Different Problem Solving Strategiesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Task Offloading Based on LSTM Prediction and Deep Reinforcement Learning for Efficient Edge Computing in IoT

Chen

Yan

et al. 2022

Future Internet

View full text Add to dashboard Cite

show abstract

“…Ruan et al [27] assumed an energy management infrastructure as a cloud model comprising three tiers, and optimized for latency using joint optimization between Stackelberg and Lyapunov-based pricing and energy demand. Thai et al [28] proposed vertical and horizontal offloading models between edge to cloud and edge to edge, and used an approximation algorithm to minimize the cost.…”

Section: Previous Workmentioning

confidence: 99%

Comparative Analysis of 5G Mobile Communication Network Architectures

et al. 2020

View full text Add to dashboard Cite

Mobile communication technology is evolving from 4G to 5G. Compared to previous generations, 5G has the capability to implement latency-critical services, such as autonomous driving, real-time AI on handheld devices and remote drone control. Multi-access Edge Computing is one of the key technologies of 5G in guaranteeing ultra-low latency aimed to support latency critical services by distributing centralized computing resources to networks edges closer to users. However, due to its high granularity of computing resources, Multi-access Edge Computing has an architectural vulnerability in that it can lead to the overloading of regional computing resources, a phenomenon called regional traffic explosion. This paper proposes an improved communication architecture called Hybrid Cloud Computing, which combines the advantages of both Centralized Cloud Computing and Multi-access Edge Computing. The performance of the proposed network architecture is evaluated by utilizing a discrete-event simulation model. Finally, the results, advantages, and disadvantages of various network architectures are discussed.

show abstract

“…However, the use of Cloud Computing (CC) for task offloading of the end-devices can generate two major issues: high transmission latency and capacity-demand mismatch, i.e., resource overprovisioning, which leads to resource and energy waste [4]. To mitigate this, the Edge Computing (EC) approach, which pushes computing capabilities at the Edge of the network, is being rapidly adopted and seems promising in terms of achieving the ambitious millisecond-scale latency required in various 5G and IoT applications [5].…”

Section: Introductionmentioning

confidence: 99%

ENERDGE: Distributed Energy-Aware Resource Allocation at the Edge

Avgeris

Spatharakis

Dechouniotis

et al. 2022

Sensors

View full text Add to dashboard Cite

Mobile applications are progressively becoming more sophisticated and complex, increasing their computational requirements. Traditional offloading approaches that use exclusively the Cloud infrastructure are now deemed unsuitable due to the inherent associated delay. Edge Computing can address most of the Cloud limitations at the cost of limited available resources. This bottleneck necessitates an efficient allocation of offloaded tasks from the mobile devices to the Edge. In this paper, we consider a task offloading setting with applications of different characteristics and requirements, and propose an optimal resource allocation framework leveraging the amalgamation of the edge resources. To balance the trade-off between retaining low total energy consumption, respecting end-to-end delay requirements and load balancing at the Edge, we additionally introduce a Markov Random Field based mechanism for the distribution of the excess workload. The proposed approach investigates a realistic scenario, including different categories of mobile applications, edge devices with different computational capabilities, and dynamic wireless conditions modeled by the dynamic behavior and mobility of the users. The framework is complemented with a prediction mechanism that facilitates the orchestration of the physical resources. The efficiency of the proposed scheme is evaluated via modeling and simulation and is shown to outperform a well-known task offloading solution, as well as a more recent one.

show abstract

Workload and Capacity Optimization for Cloud-Edge Computing Systems with Vertical and Horizontal Offloading

Cited by 60 publications

References 21 publications

Task Offloading Based on LSTM Prediction and Deep Reinforcement Learning for Efficient Edge Computing in IoT

Task Offloading Based on LSTM Prediction and Deep Reinforcement Learning for Efficient Edge Computing in IoT

Comparative Analysis of 5G Mobile Communication Network Architectures

ENERDGE: Distributed Energy-Aware Resource Allocation at the Edge

Contact Info

Product

Resources

About