Detection of Cyber-attacks using Deep Convolution in Honeyed Framework

Abstract

A network-related environment called a “honeyed framework” served to defend official network resources against harm. This framework creates a scenario that motivates the intrusive person to engage in resource-stealing activity. To recognise an unauthorised assault, this framework applied the Attack-detection-procedure. Here, we attempt to identify DoS attacks using the suggested Honeyed framework system. In order to safeguard your network from assaults, NIDS (Network Intrusion Detection System) is one of the first security solutions to make it easier to identify intrusions. In this work, we offer a system that reveals an assault while validating the defense against it. The new cyber security benchmark IoT dataset is used in this white paper to assess the most recent machine learning techniques. This work’s major goal is to develop an architecture that can foresee and stop DDOS attacks, malware, and botnet attacks using these Honeyed designs. Deep Convolution Reinforcement Neural Networks are used for network surveillance and to categories network users from potential threats (DCRNN). A two-step technique of network understanding is used to enhance the functionality of the suggested solution. DSAE (Deep Sparse Auto Encoder) is used for feature engineering challenges at the initial step of the processing process, data pre-processing. The Deep Convolution Reinforcement Neural Network learning strategy is used in the second step to facilitate categorization. The honeyed firewall and web server are then implemented, following the deployment of the honeyed framework. The DCRNN deployment is finished, and users can now be monitored and analyzed as well as data on network users collected. In this study, data from a loT environment was used to test the effectiveness of the published technique. This data included the heterogeneous datasets "IoT-23," "NetML-2020," and "LITNET-2020." With contemporary methods for network discovery, the statistical relevance of this strategy is evaluated.