Home

Introduction

Computer networks are the most common, and natural, means of use for work and commu- nication by the masses leading to a huge volume of data transactions. The volume of data transmitted has increased significantly in recent years, containing sensitive data and confi- dential information that has been successively used for illicit purposes.
In order to mitigate the invasion of privacy of data hosted in organizations’ information systems, several security solutions such as IDS have been implemented. These solutions, when implemented, need to go through a set of tests in order to guarantee their robustness and ability to act in the various intrusion situations. It is in this problematic that this work focuses where through, the generation of datasets composed of network packets of normal and abnormal traffic (attacks) in pcap format intended to test the robustness of the imple- mented solutions. Since it is intended to fill the currently gap of reduced existence of test datasets representing approximate traffic to real data. This construction is based on real data, pre-captured in the network under test, to be later studied statistically in order to find a beha- vior/use pattern to adapt the resulting data to the reality in question. This approach focuses mainly on the reality of the captured data, thus allowing to produce datasets with a similar degree of similarity to reality.
In addition to the technical approach in pcap format, this work also intends to test more specific mechanisms like IDS produced in the scope of research works. For this you can create datasets in plain text format, where the fields that compose them, as well as their order, are fully customizable as required.