Edge enhanced deep learning system for IoT edge device security analytics

Abstract

The processing of locally harvested data at the physically accessible edge devices opens a new avenue of security threats for edge enhanced analytics. Cryptographic algorithms are used to secure the data being processed on the edge device. However, the implementation weakness of the algorithms on the edge devices can lead to side‐channel attack vulnerability, which is exacerbated with the application of machine‐learning techniques. This research proposes a deep learning‐based system integrated at the edge device to identify the side‐channel leakages. To design such a deep learning‐based system, one of the challenges is formulating the suitable attack model for the underlying target algorithm. Based on the previous findings, three machine learning‐based side‐channel attack models are curated and investigated for the edge device security evaluations. As a test case, the standard elliptic‐curve cryptographic algorithm is selected. Moreover, quantitative analysis is provided for the best attack model selection using standard machine‐learning evaluation metrics. A comparative analysis is performed on the raw unaligned data samples and reduced feature‐engineered samples using edge enhanced security analytics. The investigation concludes that the vulnerable algorithm implementation can lead to the secret key recovery from the edge device, with 96% accuracy, using a neural‐network‐based algorithm to analyse side‐channel attacks.

Systematic Literature Review of EM-SCA Attacks on Encryption

“…These template attacks are often quite successful against hardware-implemented advanced encryption algorithms, as shown by Bukasa et al [199]. Machine learning-based attacks have been applied as a cutting-edge method that involves utilizing machine learning algorithms to examine the EM emissions produced by a device as it executes cryptographic operations [80,100,113,117]. These machine learning-based attacks can be particularly successful since they automatically recognize EM emission patterns that signify the use of the encryption key.…”

“…They benefit from the information that unintentionally leaked from a device that was designed to prevent tampering. Mukhtar et al [100] used a side-channel EM leakage trace collected from an FPGA to conduct the Elliptic curve scalar multiplication (ECSM). The EM probes are employed to collect the leakages, which are then recorded properly to create a leakage dataset for further side-channel investigation.…”

“…Additionally, the SEMA method has revealed that two additional ECC-based algorithms, Elliptic Curve-based Diffie Hellman (ECDH) and Elliptic Curve-based Digital Signature Algorithm (ECDSA), are susceptible to EM side-channel attacks [101]. Many IoT and mobile devices frequently use ECC algorithms to protect the data, since of their minimal computing overhead [2,100]. This reveals that such devices can be analyzed through EM side channels to obtain access to data that is encrypted.…”

Edge intelligence‐enabled cyber‐physical systems

“…This vulnerability will become more and more serious. Mukhtar et al 11 proposed a deep learning‐based evaluation system integrated at the edge device to detect side channel leaks when edge devices deploy new security algorithms for updating. The model is trained in the cloud and deployed at the edge after the training is completed.…”