Loading...

Workflow Ensemble Scheduling in Edge-Based Infrastructures using Game Theory

Parto, Hossein | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57116 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Fazli, Mohammad Amin
  7. Abstract:
  8. With the expansion of the Internet of Things and the significant increase in latencysensitive applications, cloud computing is facing serious challenges. In order to solve the challenges of cloud computing and respond to the needs of delay-sensitive applications, the concept of edge computing is proposed. In edge-based infrastructures, computing resources are located near the network edge and end devices. These infrastructures can better serve latency and security sensitive applications that require low processing and high interaction with end devices. However, achieving the benefits of edge-based infrastructure requires optimal resource management. One of the effective steps for optimal use of resources is proper timing and assigning appropriate resources to each task. In recent years, many scheduling methods have been presented, most of these methods are designed for cloud computing environments, and the challenges and sensitivities of edge-based infrastructures have not been considered in them. In this research, we intend to provide an optimal scheduling method for workflows in edge-based infrastructures by accurately understanding the challenges and sensitivities of edge-based infrastructures and workflows. Workflows are used to model relationships between tasks and represent a more realistic form of software. The proposed solution includes a two-step algorithm. In the first step, the tasks are placed in the priority queue with the new criteria that we have provided based on the specific conditions of the workflows. In the second step, the problem will be modeled as a stackerberg game. In this game, the tasks of the parent will be the role of the leader and the tasks of the child will be the role of the follower. The child makes the optimal decision based on the parent’s decision. A dynamic algorithm is used to find the equilibrium point of this game. Simulation in different conditions and with different criteria shows that the presented method has better performance in terms of cost and execution time than other methods
  9. Keywords:
  10. Task Scheduling Algorithm ; Edge Computing ; Game Theory ; Workflow ; Cloud Computing

 Digital Object List