Loading...

Distributed Latency Aware Virtual Machine Placement for Multiple Applications in Mobile Edge Clouds

Hashemi, Boshra Sadat | 2019

504 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52291 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Goudarzi, Maziar
  7. Abstract:
  8. In the near future, we will see a new generation of cellular networks called 5G. Due to the growing number of users and applications, cloud computing does not provide users needs. Edge servers can be used at radio base stations for this purpose. Edge server is a limited processing unit that allows users to perform their processing operations as close as possible. Therefore, the distance between the user and the service will be reduced and as a result the response time is greatly decreased.Cloud-based services are supported by specific software that are deployed in a virtual machine. The number of virtual machines of different applications is fewer than the number of edge servers in the network. Failure to place these virtual machines can lead to an unacceptable increase in response time to a single request. The decision-making algorithms for placement are mainly centralized and are not scalable. As the number of edge servers, applications and users increase, the running time of algorithms become unacceptable. The problem addressed in this research is to place virtual machines of users' applications on a set of edge servers by a distributed algorithm to obtain better scalability than the centralized method, while reduces the mean and maximum response time to a single request. The results of the simulation show that the running time of the distributed algorithm is linearly increasing with the increasing number of edge servers, users and applications. At the same time, it reduces the average response time by 30% and the maximum response time by 38% compared to the centralized algorithm
  9. Keywords:
  10. 5th Generation Cellular Network ; Mobile Edge Computing ; Average Response Time ; Distributed Algorithm ; Virtual Machine ; Scalability

 Digital Object List

 Bookmark

No TOC