Leveraging Timing Side-Channel Information and Machine Learning for IoT Security
Digest of Technical Papers - IEEE International Conference on Consumer Electronics
In the recent era, adoption of IoT technology can be seen in nearly every other field. However, with its fast growth rate, the IoT brings several advanced security challenges. To alleviate these problems, we propose an inventive framework that could be introduced into the IoT package to gather side-channel information (execution timing) for inconsistency and attack location. We observed that when there is an attack on the IoT device, the applications running on the device will take more time to execute as the resources such as memory, I/O, CPU will get exhausted. Execution time collected from IoT devices under normal and attack mode will be converted to digital signals, from which features will be extracted and used to build a fingerprint. The support vector machine technique will be adopted to help identify the operating status of the IoT device.
IoT, Machine learning, side-channel-attack
Kratika Sahu, Rasika Kshirsagar, Surbhi Vasudeva, Taghreed Alzahrani, and Nima Karimian. "Leveraging Timing Side-Channel Information and Machine Learning for IoT Security" Digest of Technical Papers - IEEE International Conference on Consumer Electronics (2021). https://doi.org/10.1109/ICCE50685.2021.9427585