Transformers : Intrusion Detection Data In Disguise

Boorman, James and Prince, Daniel and Green, Benjamin (2020) Transformers : Intrusion Detection Data In Disguise. In: 3rd International Workshop on Attacks and Defences for Internet-of-Things, 2020-09-19 - 2020-09-19, Online.

Text (ADIoT-camera-ready)
main.pdf - Accepted Version
Download (253kB)

Abstract

IoT cyber security deficiencies are an increasing concern for users, operators, and developers. With no immediate and holistic devicelevel fixes in sight, alternative wraparound defensive measures are required. Intrusion Detection Systems (IDS) present one such option, and represent an active field of research within the IoT space. IoT environments offer rich contextual and situational information from their interaction with the physical processes they control, which may be of use to such IDS. This paper uses a comprehensive analysis of the current stateof-the-art in context and situationally aware IoT IDS to define the often misunderstood concepts of context and situational awareness in relation to their use within IoT IDS. Building on this, a unified approach to transforming and exploiting such a rich additional data set is proposed to enhance the efficacy of current IDS approaches.