文章基本信息

标题：CNN-based Security Authentication for Wireless Multimedia Devices
本地全文：下载
作者：Gautham SK ; Anjan K Koundinya
期刊名称：International Journal of Wireless and Microwave Technologies(IJWMT)
印刷版ISSN：2076-1449
电子版ISSN：2076-9539
出版年度：2021
卷号：11
期号：4
页码：1-10
DOI：10.5815/ijwmt.2021.04.01
语种：English
出版社：MECS Publisher
摘要：Security is a major concern for wireless multimedia networks because of their role in providing various services. Traditional security techniques have inadequacies in identifying emerging security threats and also lacks in computing efficiency. Furthermore, conventional upper-layer authentication doesn’t provide any protection for physical layer, thus leading to leakage of privacy data. Keep these issues in mind, the paper has envisioned an artificial intelligence-based security authentication system that is lightweight, adaptive and doesn’t require any explicit programming. The neural network is built on convolutional filters which explore the data and learns the features or characteristic of the data. With this learned feature, the model will be able to recognize whether a wireless multimedia device present in a network is legitimate or not. Experimental analysis and validation have been performed on the trained model and ensure that the authentication of wireless multimedia devices can be achieved and also ensuring lightweight authentication system, which ensures less computation needs. The different neural model is also trained using gaussian noise of different standard deviation so that it can be used in a practical scenario like smart industry etc.
关键词：Wireless Multimedia Networks;Security;Neural Networks;Gaussian Noise;Convolutional Layer