Trend analysis of Snort alarms

In the present world, with increasing Internet availability, there is a constant threat about data, identity and other confidential information being stolen by online intruders. Delivered technologies provide great services and opportunities to increase productivity for everyone, but they do not protect users and businesses on the appropriate level. The network security became a crucial element of every company nowadays and it stopped being just an unnecessary, expensive "add-on" as it used to be. Snort - lightweight Intrusion Detection System filled a niche for lacking part of the whole security concept. Firewalls and other protection's mechanisms are still very popular but in order to provide defence-in-depth approach, the IDS is a "must" condition. The phenomenon of Snort lies mainly in two aspects. Firstly, fully customized, simply rules language, which is capable to work with both, small private networks or large scale, complex infrastructure. Secondly open source structure, which can be used and developed by everyone, makes it a unique, high professional security tool. However Snort is freely available, the implementation of IDS and further analysis of Snort alerts is not a simple task and requires highly educated people working with many extemal, analysis applications. The purpose of this paper is to provide a comprehensive set of information about current and future attack trends. Firstly by analysis, existing reports and literature about present attack patterns and trends. On the other hand by physically collection IDS alerts and further analysis of gathered samples, in order to find the future pattern of network attacks. The collection was made on two, separate, completely different topologies in order to get the most extensive point of research.School of Computin