SSENet-2014 Dataset: A Dataset for Detection of Multiconnection Attacks

Multiconnection attacks such as DoS, probe, flooding, etc., have become common and attackers have come out with sophisticated techniques as well as tools to launch variants of such attacks. This growing amount of attack and sophistication has given rise to the increasing need of efficient detection algorithm. To test and compare the performances of the proposed detection algorithms, benchmark datasets are required to represent the dynamic nature of the network. Though certain benchmark datasets are available, most datasets are either synthetic or contains suppressed information. In this paper, we introduce SSENet-2014 dataset which is generated in a real network environment. The attacks were generated using attack tools while carrying out normal activities. The description of the SSENet-2014 dataset is given. Then, a comparison is carried out with the most popular intrusion detection dataset, 10% KDD Cup 99. Two clustering approaches of K Means and Self Organizing Map (SOM) have been used in our experiments. Box plot is used to analyze the attributes of the two datasets. The results confirm the variability existing in the attribute values of 10% KDD Cup 99 and SSENet-2014 dataset. Also, it can be seen that SSENet-2014 dataset generated from a real network varies considerably from 10% KDD Cup 99 which is generated from simulated traffic.