Malware Threats and Detection for Industrial Mobile-IoT Networks

Abstract

Industrial IoT networks deploy heterogeneous IoT devices to meet a wide range of user requirements. These devices are usually pooled from private or public IoT cloud providers. A significant number of IoT cloud providers integrate smartphones to overcome the latency of IoT devices and low computational power problems. However, the integration of mobile devices with industrial IoT networks exposes the IoT devices to significant malware threats. Mobile malware is the highest threat to the security of IoT data, user’s personal information, identity, and corporate/financial information. This paper analyzes the efforts regarding malware threats aimed at the devices deployed in industrial mobile-IoT networks and related detection techniques. We considered static, dynamic, and hybrid detection analysis. In this performance analysis, we compared static, dynamic, and hybrid analyses on the basis of data set, feature extraction techniques, feature selection techniques, detection methods, and the accuracy achieved by these methods. Therefore, we identify suspicious API calls, system calls, and the permissions that are extracted and selected as features to detect mobile malware. This will assist application developers in the safe use of APIs when developing applications for industrial IoT networks.

Keywords

• Industrial mobile IoT
• threats
• malware
• detection method
• machine learning