查看文章

The existing intrusion detection systems (IDS) have found it very demanding to detect growing cyber-threats due to the voluminous network traffic data with increasing Internet of Things (IoT) devices. Moreover, security attacks, on the other hand, tend to be unpredictable. There are significant challenges in developing adaptive and strong IDS for IoT to avoid false warnings and assure high detection efficiency against attacks, especially as botnet attacks become more ubiquitous. Motivated by these facts, in this paper, different types of botnet attacks have been studied and how they are more conveniently launched with open and vulnerable IoT devices. Then, the growing trend of deep learning (DL) techniques is being studied extensively for their ability to detect botnet attacks by learning from time series data specifically in the IoT environment. Hackers are exploiting the Internet of Things (IoT), creating millions of new …