Response Time Improvement Using Service Migration in Mobile Edge Computing Considering User Mobility

Please enable javascript in your browser.

Nejati, Amir Reza | 2018

753 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 51371 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Goudarzi, Maziar
Abstract:
With the emergence of the 5th generation cellular networks, response time is becoming increasingly more important. One of the solutions to satisfy this requirement is using mobile edge computing. MEC is bringing the cloud computing features to the edge of the network and reduces services response time by reducing the user distance to her services. But for this solution to be effective, we need to migrate user services based on their movements. The main problem in migrating services is predicting user route and do the migration based on these predictions. We must do this procedure in a way that minimizes the service outage. In this research, we predict the user route with the historical trajectories dataset and predict the user time of entrance to the next cell with live trajectories dataset. The results based on a real trajectories dataset shows that we can achieve more than 80% accuracy in both predictions within about 50 nanoseconds that is a lot faster than previous works
Keywords:
Mobile Edge Computing ; Response Time ; Fifth Generation of Mobile Networks ; Service Migration ; Mobile User ; Route Prediction

No TOC