Q-Learning Algorithm Based Topology Control of Power Line Communication Networks

• Published in: 2020 IEEE 11th International Conference on Software Engineering and Service Science (ICSESS) pp. 347 - 350
• Main Authors: Chen, Wenbin, Zheng, Libin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16.10.2020
• Subjects: component; Link Reliability; Network topology; Power Line Communication; Power line communications; Prediction algorithms; Q-Learning; Reliability; Testing; Time series analysis; Topology; Topology Control
• ISBN: 1728165784; 9781728165783
• ISSN: 2327-0594
• DOI: 10.1109/ICSESS49938.2020.9237707

As link reliability is increasingly important in power line communications (PLC), the issue of link quality prediction has already become necessary with real-time accuracy to guaranty communication services in smart power grid. However, the superposition of nonlinear and non-stationary sequences expressed by signal to noise ratio time series of PLC link quality is the main factor that affects the performance when the PLC network conducts topology control operation. Aimed to solve this problem, this article proposes a Q-Learning algorithm enabled topology control scheme in PLC networks. In this proposed approach, the link reliability prediction model based in Q-Learning is established, which use the received signal strength information between adjacent nodes to determine the connection state between adjacent nodes. Moreover, it makes full advantage of the PLC link reliability prediction result to conduct topology control operation to achieve less packets loss and higher efficiency. Testing results show that the proposed approach is able to improved services supporting ability of PLC networks.