Cyber Security in IOT Networks (Using Hybrid Approach)

Abstract

Cyber security refers to the act of protecting systems, networks, and data form unauthorized access. There are great chances of hacking of data in today’s connected world where everyone is connected with internet through numerous devices like laptop, mobile phone, tablet, IOT devices. The use of IOT devices is increasing exponentially due to ease of use and less overhead. The area of implementation ranges from healthcare, medical, smart homes, smart cities to logistics, manufacturing, agriculture. They are offering seamless services and people feels like a magic while using these devices. But these devices are lacking strong security measures which can cause unauthorized access to these devices. They are vulnerable to various types of cyber-attacks like denial of service (DOS), ransom ware attack, malware attack, node spoofing and zero-day attack. Various intrusion detection systems are developed to detect these attacks using pre-existing data. Zero-day attack is severe among these as it cannot be detected using the previous cyber-attack data available. These are very difficult to detect as their type is unknown and traditional IDS are unable to detect them due to unavailability of their signature. So, there is a need to develop such type of systems which can detect unknown types of cyber-attacks. This research paper presents the role of different machine learning techniques in cyber-attack detection of different types in IOT networks along with comparative study. The challenges and issues are also identified and future research directions are given. A model is also proposed for zero-day attack detection.

Keywords: Cyber-attack, Internet of things, Intrusion detection system, Machine learning, Security, Zero-day attack.

View PDF