The Internet of Things (IoT) has been proliferating over the past few years,
and is expected to continue spreading quickly on a global scale.
Due to inherent characteristics, IoT devices are commonly more susceptible to cyber-attacks when compared to servers, personal computers and even tablets and smartphones.
Organizations deploying such a technology, while hosting an increasing abundance of smart devices in their wireless networks, have begun to face new related management and security challenges.
In the related research projects conducted at the IoT Security Lab, analyzing network traffic data with advanced machine learning methods is proposed in order to profile diversified IoT device types, thus facilitate organizational security applications.
The IoT research includes identification and continuous verification of connected IoT devices, anomaly detection for intrusion detection, identification of vulnerable IoT devices behind NAT in a privacy preserving approach, evaluating novel attack scenarios and attack vectors involving IoT devices, etc.
Increasing amounts of traffic data are collected from a plethora of IoT devices, models and categories, deployed across numerous labs at Ben-Gurion University (Israel) operated naturally over long periods of time.
The proposed methods, evaluated on these authentic, abundant and diverse data, may serve to facilitate modern organizations in enforcing complex security policies and better cope with IoT-inflicted threats.