The rise of Internet of Things (IoT) has envisioned smart connectivity of billions of varied devices with high computational and processing capabilities resulting in the evolution of Internet of Medical Things (IoMT). This radical increase in the digital landscape of IoMT that processes considerably large amount of valuable data is primarily a potential attack target for varied cyber adversaries. One of the most prevalent, sophisticated and new evolving cyber threats for IoT ecosystems are multifaceted malicious malwares. The authors propose a highly scalable hybrid (deep learning) DL-driven SDN-enabled framework for efficient and timely detection of sophisticated IoMT malwares. Further, the proposed framework leverages the underlying IoMT resource constrained devices without exhaustion. We employed state-of-the-art publicly available dataset for a comprehensive evaluation of the proposed mechanism. Further, standard metrics have been employed to rigorously evaluate the performance of the proposed technique. For verification purpose, we compare our proposed mechanism with our constructed hybrid DL-driven architectures comprised of state-of-the-art DL-algorithms and current benchmarks. The proposed scheme outperforms in terms of high detection accuracy and speed efficiency. We also employed 10-fold cross validation to explicitly show unbiased results.
Bibliographical noteFunding Information:
This work was supported by the European Commission , under the ASTRID and FutureTPM projects; Grant Agreements no. 786922 and 779391 , respectively.
© 2021 Elsevier B.V.
- Deep learning (DL)
- Executable malware's
- Factory-of-things (FoT)
- Hybrid deep learning architecture
- Internet of Medical Things (IoMT)
- Internet of Things (IoT)
- Software Defined Networks (SDN)