Abstract:
An acquiescent juxtaposition of wireless technologies, micro-electromechanical systems (MEMS), micro-services and the internet paved an ecosystem named Internet of Things (IoT), which ascertains link between physical objects that are reachable through the internet. The embedded technology in those objects succors them to interact with internal states or the external milieu, which in turn influences on decisions. This new connectivity, bridges the gap between physical objects and digital world to improve the quality and productivity of life, has become common and going beyond laptops and smartphones, in applications like cars, smart homes, smart cities, healthcare, retails, energy management agriculture, wearables etc. IoT connects smart objects together (through internet and intelligent sensors) using internet protocol, and make them to be read, controlled, and managed at any time at anywhere. Since this communication is in the public environment, these devices are vulnerable to attacks, and hence the security and privacy are vitiated. Detection of abnormality in propositional information must be followed by recovery action to ensure the correct semantics of the frame network. This paper focuses on building a semantic based security platform to analyze the data received from sensors using Hidden Markov Model (HMM), semantic sensor network ontology, and temporal ontology to detect the malicious attack data. The HMM is used for reasoning purpose and the label for visible states are created. The Stream Annotation Ontology is used to represent the quality of the data over the Semantic Sensor Network Ontology.
