We propose a novel speech keyword spotting mechanism which is robust against attacks exploiting the inherent vulnerabilities of an ML model in correctly classifying the spoken word.
We propose a novel program integrity verification mechanism which is inherently designed considering the scalability in a swarm setting reducing the total swarm attestation time as compared to the prior art.
We present a practical runtime remote attestation framework that segments the software and randomizes the integrity check of these segments over short random intervals leading to a novel performance-vs-security trade-off.
We investigate lightweight mechanisms for checking the integrity of the software running on IoT Devices.
To provide message authenticity in IoT, we employ a speculation procedure for predicting future message values to achieve an advantageous trade-off between the cryptographic strength and the latency in processing of the message authentication codes.
To provide message authenticity in IoT, we enable the verifier to achieve an advantageous trade-off between the cryptographic strength and the latency in processing of the message authentication codes.