Security considerations in IoT using machine learning and deep learning
S. Hemalatha, K. Jothimani, S. Anbukkarasi · 2024
Internet of Things (IoT) is used in various domains such as Industrial 4.0, smart home, social media healthcare, and smart energy. To develop the future smart home technology combines the concept of interoperability, reliability, connectivity, and machine intelligence. Availability of smart home technology makes people feel safe at home. Information related to smart appliances is stored in cloud storage. Its monitoring and scheduling can be performed with the help of cloud computing so it enables reliability of the power system. In the cloud, due to lack of security the sensitive data is accessed by hackers that lead to many issues. Before communicating the confidential information to other devices we are in need to ensure the security in all fields. Both privacy and security have significant challenges for managing smart appliances. There are several types of security attacks possible in smart appliances using IoT; these attacks include denial of service, encryption attacks, phishing, obfuscations, botnets, jamming, man-in-the-middle, eavesdropping, ransomware, invasions, brute force password attack, and other cyber threats to IoT systems. Analysing the suitable deep learning techniques will improve security. However, artificial intelligence (AI) is the science and engineering technology to make intelligent machines that work and behave like human beings. Machine learning (ML) is the subset of AI; deep learning is the subset of ML. These three techniques are interconnected with each other for managing automated activities in smart appliances and also providing security. There are several machine learning algorithms available in ML. Conventional neural network and generative adversarial networks are the most commonly used ML method. Unnecessary notifications are the major challenge in smart home appliances. In this, the architecture of IoT systems and its threats are notified. The recent research direction is to identify the safety and security of IoT using the integration of ML and DL techniques. Additionally, it analyses potential difficulties for future analysis when using security techniques in IoT.