Network security management using ontology-based mobile agents

Abstract EN

Automatic means to manage the security in moderate and large networks are of extreme importance to avoid error-prone manual techniques.

This paper paves the way to develop an automatic network security management (NSM) system that is both flexible in deciding the system’s objectives and efficient in using the valuable network bandwidth with a relatively low transmission overhead.

The required flexibility and efficiency are obtained using mobile agents (MA) to collect the required security information from various network components and devices, and using ontology to specify the required security policies in such a way understandable by the MA’s software.

A simplified NSM prototype is developed, implemented, and tested over a typical local area network to investigate the validity of the suggested ideas.

The MA travels through the network and collects the necessary information using an ontology-based security policy.

Next, it may either return back to the network administrator to let him decide and perform the suitable actions, or alternatively the MA may take the appropriate decisions.

This prototype is tested to examine its functionality and performance using a simple local area network consisting of three computers with different configurations.

The developed MA was able to understand the ontology and move around the network.

It has properly detected the components that are wrongly configured.

It should be made clear that the design is scalable and can be directly applied to more computers in a local area network or even in a wide area network.