Dataset AttackUDPModel-Training

This dataset was exploited within the scope of the REACT project (WP4/WP5) to provide a means for attack behaviour modelling for UDP flooding attacks.The work flow constituted:- To generate data related to the performance of the network under simulated malicious scenarios (i.e., simulated attacks) in Cooja. In particular, the data should be measurements of network performance metrics (e.g., throughput, response time, packet loss) and device performance metrics (e.g., energy consumption, transmitted/received packets to allow the identification of proper attack impact metrics.- Relevant and informative features are to be derived from the identified attack impact metrics so as to improve the detection capabilities of AIDSs (Anomaly-based Intrusion Detection Systems) and at the same time reducing their false-positive rate.- To generate attack behavior models based on the simulated data related to the generated network traffic and the log files of the IoT devices of simulation scenarios where massive IoT networks are under attacks. The model development will be based on software tools such as MATLAB and Python along with the Scikit-Learn library.In this context, the work flow was adhered to that resulted in the characterisation of relevant metrics for UDP flooding