A Scheme of Anomalous Detection Based on Reinforcement Learning for Load Balancing. Issue 1 (March 2020)

• Record Type: Journal Article
• Authors: Kim, Hye-Young
• Is Part Of: IOP conference series. Volume 790:Issue 1(2020)
• Publication Date: 2020-03
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/790/1/012035
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: In recent, both researchers and developers have great interests in anomalous detection. However, it is still difficult to implement a uniform framework for anomalous detection. Also, the network anomalous detection using deep learning methods has been discussed with potential limitations and interests. An anomalous detection in wireless or wired network is extremely important because it is caused by flood traffic of network and intrusion. Patterns of malicious network loads are defined, while anomalous detections is more suitable for detecting normal and anomalous network loads by means of deep learning. The important goal of these issues is to recognize the anomalous detections for better preparation against future load balancing of networks. In this paper, we propose an agent Detectbot that processes anomalous detection for load balancing based on reinforcement learning. Our simulation results show that the reinforcement learning scheme is effective for anomalous detection in load balancing.