Enhancing anomaly detection for social multimedia in SDN using glow worm-driven deep belief network

The evolution in network security within social multimedia communication depends on effective anomaly detection. One of the most essential elements of security involves identifying suspicious behavior during multimedia data transport. These abnormalities can adversely impact the network's perfo...

Full description

Saved in:
Bibliographic Details
Main Authors: S. K. Manju Bargavi, Chintan Thacker, Varsha Agarwal, Yaduvir Singh, Sneha Kashyap, Dhananjay Kumar Yadav
Format: Article
Language:English
Published: Taylor & Francis Group 2025-07-01
Series:Automatika
Online Access:https://www.tandfonline.com/doi/10.1080/00051144.2025.2476806
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The evolution in network security within social multimedia communication depends on effective anomaly detection. One of the most essential elements of security involves identifying suspicious behavior during multimedia data transport. These abnormalities can adversely impact the network's performance and reliability. Proactive anomaly detection and mitigation in social multimedia communication is rendered possible by the combination of Software-Defined Networking and advanced analytics. By identifying anomalous behavior in the transfer of social multimedia data, the technique improves security. In this paper, suggested a Glow Worm-Driven Deep Belief Network (GW-DDBN) to enhance anomaly detection, and Principal Component Analysis is used for dimensionality reduction. To apply the shortest path routing to transmit the component in the control layer. To gathered 400 real-time users’ dataset who were active on the internet for one hour and used multiple digital media platforms encompasses Facebook, Instagram, WhatsApp, and Twitter. As a result, to assess the performance of the suggested method as it relates to packet counts over some time, latency, bandwidth, and consuming energy, then compare our proposed method to the existing method concentrated on metrics including accuracy (98%), precision (97%) and recall (95%), F1 score (85.6%). The method demonstrates superior performance in identifying and addressing network irregularities.
ISSN:0005-1144
1848-3380