Trust Oriented Efficient Data Aggregation and Malevolent Node Detection in the Internet of Things Sensor Network

Sensor technologies that enable monitoring and sensing of various environmental events in the Internet of Things are driven by limited energy resources and computing constraints. To improve energy efficiency,the sensor node is required to perform data aggregation process to eliminate redundancy in the information obtained from adjacent nodes before transmitting to the local access point. However, the sensor nodes are vulnerable to many security threats, the information gathered can be falsified by malevolent nodes. Many solutions for reliable information aggregation and malevolent node detection have been introduced in the existing literature. However, all these solutions have certain limitations that do not meet the requirements of maintaining higher energy efficiency, maximum security, and low computational overhead. To this end, this paper proposes a quantitative trust model for evaluating the reliability factors of each communicating node. The model computes trust values based on the behavioral analysis of each node, and the optimal nodes with higher trust valuesare selected for the clustering and data aggregation process. Malevolent nodes are identified by a trusted entity located in the base station that uses statistical analysis for the received sensor node information. The proposed model is implemented on a numerical computing tool. The effectiveness and scope of the proposed system are assessed regarding packet delivery ratio, network overload, and energy utilization. The outcome shows that proposed model achieved 16% improvement in the packet delivery ratio, 16.6% improvement in reduction of network overhead, and 27.73% improvement regarding energy efficiency compared to existing method.