Security Threat Prediction in a Local Area Network Using Statistical Model

In today´s large and complex network scenario vulnerability scanners play a major role from security perspective by proactively identifying the known security problems or vulnerabilities that exists across a typical organizational network. Identifying vulnerabilities before they can be exploited by malicious user often helps to test, maintain, and assess the risk of the existing network. Still there are many problems with currently available state of the art vulnerability scanners like hampering system resource. One possible solution to this problem might be reducing the number of vulnerability scans, along with the quantitative approach towards different vulnerability category in order to identify which class of vulnerability should enjoy preference in the risk mitigation procedure. This paper introduces a model that predicts vulnerabilities that will occur in near future on a local area network (LAN) by using statistical measures and vulnerability history data. Two case studies have also been presented to validate the model.