9th International Conference on ICT for Sustainable Development

Authors - Zakiyabanu Malek, Himali Gajjar
Abstract - Internet apps have grown in popularity now a days. As the use of internet technology is increasing day by day security concern over internet is also increasing. For that we need network to be more secure to overcome the security concern it is mainly depending on intrusion detection systems (IDSs).Many technologies have already develop to fight against network security concern, this technology is mainly work with the help of deep learning as well as artificial intelligence techniques Deep leaning is heavily effective in IDSs as a metafield of AI. Convolutional neural networks (CNNs) use a common deep learning neural network topology to process complicated input. CNN is commonly used in intrusion detection systems (IDSs) and overcomes the limitations of classic machine learning approaches. IDSs manage security and privacy concerns using a variety of CNN-based approaches. However, to the best of our knowledge, no comprehensive assessments of IDS projects have used CNN. The primary goal of this research is to gain a better understanding of the many way and working of CNN, irregularities, and types of assaults. In a study artistically organises the major characteristics and contributions of the examined CNN-IDS approaches into several categories. These approaches are compared based on their essential components, which include the classifier method, architecture, input shape, performance, evaluated metrics, and dataset. The experimental findings of CNN-IDS research are not comparable since different datasets are used. As a consequence, an empirical experiment was conducted in this study to evaluate alternative approaches using combined datasets. In this paper we have explained some results in deep.