Tehnički vjesnik, Vol. 30 No. 6, 2023.
Izvorni znanstveni članak
https://doi.org/10.17559/TV-20230323000475
Development of Bandwidth Optimization and Limiter Software for Network Efficiency in Software-Defined Networks
Abdülkadir Çakir
; Isparta University of Applied Sciences, Faculty of Technology, Electrical and Electronics Engineering, Turkey
Enes Açikgözoğlu
orcid.org/0000-0001-7293-883X
; Isparta University of Applied Sciences, Keçiborlu Vocational School, Computer Technologies, Turkey
Sažetak
Many devices have been connected to each other and a wide platform has been formed with the development of internet technologies. The continuous expansion of this platform has revealed requirements such as single point management, accessibility, bandwidth management and efficient use of the network. Considering that software-defined networks are systematically managed by software, it is predicted that they will meet the determined network requirements more easily. In this study, software was developed that limits network traffic on a client basis by optimizing the bandwidth of clients in software-defined networks. In the proposed study, a unique dataset was created by taking the last year's data from the university network for bandwidth optimization. In order to determine the optimum client-based bandwidth, the dataset is clustered with the K-means algorithm. The instant data coming from the live network is transferred to the software as client and the cluster to be transferred is calculated. Web-based limitation software performs network traffic limitation by including the clients in the optimum cluster according to the cluster information coming from the dataset instantaneously. A virtual network was designed for the implementation of the web-based software and tests were carried out on this network. Efficient use of the network is aimed by allocating bandwidth according to clusters created especially in multi-user, heavy-traffic networks. In addition, client-based DDoS attack detection is also carried out thanks to the network data collected instantly.
Ključne riječi
intrusion clustering; intrusion detection; K-means; SDN
Hrčak ID:
309226
URI
Datum izdavanja:
25.10.2023.
Posjeta: 267 *