Home // CYBER 2023, The Eighth International Conference on Cyber-Technologies and Cyber-Systems // View article


Bespoke Sequence of Transformations for an Enhanced Entropic Wavelet Energy Spectrum Discernment for Higher Efficacy Detection of Metamorphic Malware

Authors:
Steve Chan

Keywords: Industrial Control Systems; Distributed Control Systems; Operational Technology; Condition Monitoring Paradigm; Industrial Internet of Things; Metamorphic malware.

Abstract:
A Robust Convex Relaxation (RCR) Long Short- Term Memory (LSTM) Deep Learning Neural Network (DLNN) can provide enhanced Entropic Wavelet Energy Spectrum (EWES) discernment regarding the potential use of packers, crypters, and protectors (it has been found that compressed or encrypted files have greater entropy values), which can be indicative of Metamorphic Malware (MM). The RCR-LSTM DLNN facilitates a more robust Recurrent Neural Network (RNN) to Feedforward Neural Network (FNN) progression via a bespoke Nonnegative Matrix Factorization (NMF) to Multiresolution Matrix Factorization (MMF) to Continuous Wavelet Transform (CWT) Sequence of Transformations (SOT). Preliminary experimentation pertaining to the RCR-LSTM DLNN framework indicates potential higher efficacy for an enhanced EWES discernment than traditional Machine Learning (ML) and DLNN methods. The potential impact includes the greater use of Industrial Internet of Things (IIOT) sensors, which have been beset by MM, for Industrial Control Systems (ICS), among others.

Pages: 46 to 52

Copyright: Copyright (c) IARIA, 2023

Publication date: September 25, 2023

Published in: conference

ISSN: 2519-8599

ISBN: 978-1-68558-113-8

Location: Porto, Portugal

Dates: from September 25, 2023 to September 29, 2023