A Study on Task Offloading Scheme Using Deep Q-Network in Edge-Cloud Computing Environment

Journal of Advanced Technology Research, Vol. 7, No. 1, pp. 5-13, Jun. 2022
10.11111/JATR.2022.7.1.005, Full Text:
Keywords: Deep Q-Network, Edge-Cloud computing, Resource Allocation, Markov Process, IoT, 5G network
Abstract

With the recent development of intelligent IoT applications and 5G network services, a new paradigm has led to the emergence of edge-cloud computing. Depending on resource demand, IoT applications must offload to the edge or cloud to perform computationally complex and low latency tasks. Therefore, intelligent resource management techniques are needed for Edge-Cloud computing system to allocate resources more efficiently. In this paper, we aim to optimize task offloading and resource allocation under delay constraint to maximize resource utilization and minimize offloading rejection in Edge-Cloud computing system. We apply deep reinforcement learning to consider optimal decisions for task offloading and resource allocation. For optimal task offloading, we formulate the optimization problem as a Markov Decision Process (MDP), and then utilize Deep Q-Network (DQN) to find the optimal policy. And we prove that the proposed technique achieves better performance than the conventional heuristic approach in simulation results with maximizing resource utilization, with minimum cost.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
H. Lim, I. Ullah and Y. Han, "A Study on Task Offloading Scheme Using Deep Q-Network in Edge-Cloud Computing Environment," Journal of Advanced Technology Research, vol. 7, no. 1, pp. 5-13, 2022. DOI: 10.11111/JATR.2022.7.1.005.

[ACM Style]
Hyun-Kyo Lim, Ihsan Ullah, and Youn-Hee Han. 2022. A Study on Task Offloading Scheme Using Deep Q-Network in Edge-Cloud Computing Environment. Journal of Advanced Technology Research, 7, 1, (2022), 5-13. DOI: 10.11111/JATR.2022.7.1.005.