Improving communication precision of IoT through behavior-based learning in smart city environment

Abstract

Internet of Things (IoT) is an emergent heterogeneous communication platform that provides ubiquitous access to resources and meets the user demands. Developing smart city paradigm employs this communication platform for providing services to the users and granting distributed resource sharing. Mitigating adversary device selection in this platform prevents irrelevant data exchange and eases information access and exchange. External and internal communication hindering factors needs to be addressed to sustain uninterrupted communication. To improve the precision in communication security, an observation-based security system is modeled in this manuscript. This security system observes the local and global behavioral change in IoT device communication. The local and global behavioral changes are defined using device attributes and behavior modeling. The device and service provider input observations are processed by a neural network-based learning scheme to identify errors in the resource access. The communicating users and IoT devices are secured by selecting reputed service providers and data sources to improve the distributed resource utilization in IoT based smart city. The performance of the proposed scheme balances between security and resource utilization requirements of the users by reducing response loss and non-reputed device selection.