Title: CRAQL: a novel clustering-based resource allocation using the Q-learning in fog environment

Authors: Chanchal Ahlawat; Rajalakshmi Krishnamurthi

Addresses: Department of Computer Science and Engineering, Jaypee Institute of Information Technology, Noida, India ' Department of Computer Science and Engineering, Jaypee Institute of Information Technology, Noida, India

Abstract: Fog computing is an emerging paradigm that provides services near the end-user. The tremendous increase in IoT devices and big data leads to complexity in fog resource allocation. Inefficient resource allocation can lead to resource starvation and unable to complete the task assignment within a specific time. Hence, to enhance the efficiency of the fog resources, it is critical to perform proper resource allocation. This work targets to provide the solution to the resource allocation problem with a novel clustering-based resource allocation using the Q-learning (CRAQL) model. For this purpose, the problem is defined as a decision-making problem and formulated as Markov decision process (MDP). Next, to find the optimal resource, an enhanced optimal resource allocation (EORA) algorithm is proposed and detailed study is performed to analyse the impact of various performance parameters. Simulation results show the comparison of the EORA versus conventional Brute force method by varying the performance parameters such as learning rate and number of trials. The experimental results exhibit optimal solutions with significant improvement in learning rate at an average probability of 0.5 within limited epochs.

Keywords: internet of things; IoT; fog computing; reinforcement learning; resource allocation; Q-learning; Markov decision process; MDP.

DOI: 10.1504/IJCC.2024.139598

International Journal of Cloud Computing, 2024 Vol.13 No.3, pp.243 - 266

Received: 03 May 2022
Accepted: 20 Mar 2023
Published online: 04 Jul 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article