Search

The expansion of the Internet of Things (IoT) paradigm has brought with it the challenge of promptly detecting and evaluating attacks against the systems coexisting in it. One of the most recurrent methods used by cybercriminals is to exploit the vulnerabilities found in communication protocols, which can lead to them accessing, altering, and making data inaccessible and even bringing down a device or whole infrastructure. In the case of the IoT, the Message Queuing Telemetry Transport (MQTT) protocol is one of the most-used ones due to its lightness, allowing resource-constrained devices to communicate with each other. Improving its effectiveness, a lighter version of this protocol, namely MQTT for Sensor Networks (MQTT-SN), was especially designed for embedded devices on non-TCP/IP networks. Taking into account the importance of these protocols, together with the significance that security has when it comes to protecting the high-sensitivity data exchanged in IoT networks, this paper presents an exhaustive assessment of the MQTT-SN protocol and describes its shortcomings. In order to do so, seven different highly heterogeneous attacks were designed and tested, evaluating the different security impacts that they can have on a real MQTT-SN network and its performance. Each one of them was compared with a non-attacked implemented reference scenario, which allowed the comparison of an attacked system with that of a system without attacks. Finally, using the knowledge extracted from this evaluation, a threat detector is proposed that can be deployed in an IoT environment and detect previously unmodeled attacks. [ABSTRACT FROM AUTHOR]

The increase in the number of cybersecurity incidents in which internet of things (IoT) devices are involved has called for an improvement in the field of computer forensics, which needs to provide techniques in order to perform complete and efficient investigations in this new environment. With the aim of doing so, new devices and systems are being studied in order to offer guidelines for investigators on how to examine them. This papers follows this approach and presents a forensic analysis of the non-volatile memory of Windows 10 IoT Core. It details how the investigation should be performed and highlights the relevant information that can be extracted from storage. In addition, a tool for the automation of the retrieval of the pieces of evidence detected is provided. [ABSTRACT FROM AUTHOR]