DEVELOPMENT OF INTRUSION DETECTION SYSTEM FOR THREAT IDENTIFICATION OF INTERNET OF THINGS using ANSVN

Abstract: The Internet of Things (IoT) is an ever-growing network of smart objects. It refers to the physical objects which are capable of exchanging data with other physical objects. The IoT introduces various services, currently most of the day to day life depends on its available and reliable activities. Each and every single device and sensor in the IoT signifies a possible risk. The major challenges in IoT are Vulnerability, Trust and Data integrity, Data collection, Protection and Privacy. Therefore, the challenge of implementing threat identification in the IoT network is essential. The IoT network is secured with encryption and authentication, but it cannot be protected against cyber-attacks. Therefore, the Intrusion Detection System (IDS) for threat identification is needed to secure communication of IoT networks. To overcome these challenges we propose an Artificial Neural Support Vector Network (ANSVN) to identify these threats by multi-level scanning. The ANSVN algorithm is trained using internet packet traces, and then it may be deployed in the networks to identify Network threats and intrusions of IoT. This algorithm is fully based on neural fuzzy artificial intelligence working methodology, so it can identify the position information of node and its neighbour node to identify predefined wormhole attacks in the IoT and also identify malicious nodes of IoT by using Received Signal Strength Indicator (RSSI). We have tested the feasibility of the ANSVN algorithm using Network Simulators (NS2 & NS3). The test result shows that this approach act against sophisticated attack by improving accuracy, precision rate and reduce the false positive rate and keep guard data integrity, confidentiality and availability of IoT.

Keywords: ANSVN, IoT, RSSI

For full text visit

http://ijtcse.com/2017/10/09/december-2017-published-papers/