A Deep Learning Algorithm to Minimize Cyber-Security Attacks for Small Enterprises
Enock Mudau, Topside Ehleketani Mathonsi, Tshimangadzo Mavin Tshilongamulenzhe · 2024
As cybercriminals refine their tactics and develop novel attack vectors, small enterprises remain a prime target due to their limited resources and susceptibility to recent attack signatures. Despite their significant contribution to the economy, small businesses often lack the robust security measures and expertise employed by their larger counterparts. This vulnerability leaves them exposed to a growing array of cyber threats, including data breaches, ransomware attacks, and network infiltration. To address this critical cybersecurity gap, this research introduces a novel deep learning-based intrusion detection algorithm (DL-IDS) specifically designed to safeguard the networks of small enterprises. This algorithm continuously monitors network traffic, scrutinizing all unicast traffic for signs of malicious activity. Unlike conventional signature-based detection methods, the DL-IDS leverages its deep learning capabilities to extract meaningful patterns from vast amounts of raw data, enabling the detection of sophisticated attacks that evade traditional signature-based systems. Operating in a read-only mode, the DL-IDS unobtrusively sniffs the firewall's internal interface, analyzing network traffic without disrupting normal operations. Upon detecting suspicious or malicious activity, the DL-IDS triggers an alarm, alerting the hybrid-based intrusion detection system management server via a dedicated read/write network interface. This prompt notification empowers network administrators to take immediate action to mitigate the threat and protect sensitive data. The proposed DL-IDS offers several advantages over traditional intrusion detection systems namely, adaptability, accuracy, comprehensiveness, efficiency, and scalability. By implementing the DL-IDS, small enterprises can significantly enhance their cybersecurity posture, safeguarding their valuable data assets and protecting their operations from the ever-evolving threat landscape.