Ohmn: Android Malware Detection Method using Opcode Highway Memory Network

Abstract

In recent times, there has been a notable trend among Android malware to prioritise fast replication, hence expediting the acquisition of critical data. Adversaries have the potential to introduce hazardous modifications to code via the act of replicating it. In this research, we propose a deep learning-based network"Opcode highway memory network" (OHMN) that may detect variants of malware that are connected to one another without needing the analyst to have a background in mathematics or methodology.Input data may be standardised using sequence patch normalisation, and features can be discovered with truth interference fuzzy clustering by employing software birthmarking of malcode sequences. Once malware samples have been grouped together, the data is passed to the OHMN process, which searches for various types of malware. Malware detection on Android has been proved to work successfully in both experimental and observational settings. The Mal Radar dataset, an android malware dataset retrieved from zenodo in a python environment, was used for these tests.   The simulation results demonstrated that the suggested method outperformed the state-of-the-art technique.