When Electromagnetic Signals Reveal Obfuscated Malware: Deep and Machine Learning Use cases

The Internet of Things (IoT) is constituted of devices that are expo-nentially growing in number and in complexity. They use plentiful customized firmware and hardware, ignoring potential security issues, which make them a perfect victim for cybercriminals, especially malware authors.
We will describe a new usage of side channel information to identify threats that are targeting the device. Using our approach, a malware analyst is able to accuracy know about malware type and identity, even in the presence of obfuscation techniques which may avoid static or symbolic binary analysis. We captured 100,000 leakage traces from an IoT device infected by a miscellaneous and representative in-the-wild malware samples and realistic benign activity. Our technique does not need to modify the target device. Thus, it can be deployed independently from the resources available without any overhead. Moreover, our approach has the advantage that it can hardly be detected and evaded by the malware authors.
In our experiments, we were able to classify three generic malware types (and one benign class) with an accuracy of 99.82%. Even more, we show that our solution permits to classify altered malware samples with unseen obfuscation techniques during the training phase, and to determine what kind of obfuscations were applied to the binary, which makes our approach particularly useful for malware analysts.