Mapping Security Vulnerabilities and Malware Threats in Android-Based Iot Ecosystems: A Systematic investigation

Abstract

The convergence of the Android operating system with the Internet of Things (IoT) has produced a vast and heterogeneous ecosystem of smartphones, wearables, smart-home hubs, connected cameras, automotive head units, and industrial edge controllers that all share a common software lineage. This convergence has amplified both the reach and the consequences of cyberattacks, because vulnerabilities inherited from the mobile platform now extend into physical environments, critical infrastructure, and always-on sensor networks. This paper undertakes a systematic investigation of the security challenges and malware threats that specifically affect Android-based IoT devices. Drawing on recent peer-reviewed literature spanning static, dynamic, network-traffic, graph-based, and explainable deep learning approaches, the study first synthesizes the state of the art in Android and IoT malware research, highlighting recurring methodological patterns as well as unresolved limitations such as poor cross-device generalization, concept drift, adversarial fragility, and the scarcity of representative labelled data. Building on this synthesis, the paper develops a structured taxonomy that organizes the threat landscape into five interrelated dimensions: platform and firmware vulnerabilities, network and communication-layer threats, malware families that specifically target constrained IoT endpoints, companion-application and data-privacy risks, and adversarial threats aimed at the AI-based defences themselves. The paper further reviews the public datasets and benchmarks that underpin experimental work in this domain and discusses their coverage gaps. Finally, informed by the identified challenges, a conceptual deep learning-oriented detection and security-enhancement outline is presented, integrating multi-source feature fusion, lightweight on-device inference, federated collaborative learning, adversarial robustness, and explainability as design requirements for future defence systems. The investigation concludes that securing Android-based IoT devices requires detection frameworks that are simultaneously resource-aware, privacy-preserving, generalizable across device families, and resilient to evasion, and it outlines the research directions most likely to close the gap between laboratory performance and real-world deployment.