Detecting File based and Network (BGP) Based Anomalies Using Machine Learning for Enhanced Security

Abstract

The Border Gateway Protocol (BGP) serves as the center of global web routing; however, BGP's reliance upon trust and lack of solid authentication tools make it prone to multiple security threats such as path hijacking, prefix leaks, and ransomware-based events. Typical anomaly finding techniques, dependent on fixed rule systems or small datasets, frequently do not change to complex, changing dangers. For these shortcomings, cybersecurity is improved by way of a scalable, machine learning framework integrating real-time BGP monitoring with anomaly detection through analyzing Portable Executable (PE) structures with URL patterns. This method uses thorough analysis of Portable Executable (PE) forms and URL styles to spot oddities suggesting harmful actions. By thoroughly analyzing file signatures associated with malware, along with detecting suspicious URL behaviours, the proposed system greatly strengthens, in effect, threat detection capabilities. This automatic irregularity finding system seeks to improve the safety as well as the strength of worldwide web data flow. It does so via actively lessening many BGP-based safety problems in addition to facing new online risks.