Research on access strategy of autonomous scheduling service of communication polling system based on DQN technology

Yujie Xue; Min He

doi:10.1117/12.2635941

5 May 2022 Research on access strategy of autonomous scheduling service of communication polling system based on DQN technology

Yujie Xue, Min He

Author Affiliations +

Proceedings Volume 12245, International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022); 122450J (2022) https://doi.org/10.1117/12.2635941
Event: International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022), 2022, Sanya, China

Abstract

The traditional polling system in Internet of Things (IoT) communication suffers from low fairness under high load, and difficulty in adjusting the scheduling strategy according to the network state, and it is difficult to achieve the desired control effect by the traditional polling method. To overcome these problems, we propose a polling system based on Deep Q-Network (DQN) adaptive network load by combining deep reinforcement learning. In addition, the deep reinforcement learning approach is used to solve the high-dimensional complex problems, which are difficult to solve in the polling system model. The system is simulated for experiments, the queue length at polling epoch and mean waiting time of site information of the polling system were counted, to verify the accuracy of the method in learning to control the service strategy of each site. And a comparison experiment was conducted, demonstrates the system's superiority in controlling queue message latency and service fairness.

Citation Download Citation

Yujie Xue and Min He "Research on access strategy of autonomous scheduling service of communication polling system based on DQN technology", Proc. SPIE 12245, International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022), 122450J (5 May 2022); https://doi.org/10.1117/12.2635941

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available