Privacy Preserving Data Mining Approach for IoT based WSN in Smart City

Abstract

Wireless Sensor Network (WSN) is one of the most fundamental technologies of Internet of Things (IoT). Various IoT devices are connected to the internet by making use of WSN composed of different sensor nodes and actuators, where these sensor nodes collaborate and accomplish their tasks dynamically. The main objective of deploying WSN-based applications is to make high precision real-time observations, and it is extremely challenging because of the limited computing power of the sensors operating under constrained environments, resource constraints like energy, computation speed, bandwidth and memory, huge volume of high speed, heterogeneous and fast-changing WSN data. These challenges encouraged the researchers to concentrate deeper on exploring data mining techniques to extract the required information from the fast-changing sensor data in WSN and thereby efficiently handle the massive data generated by the WSNs. The increasing need of data mining techniques for WSN has inspired us to propose a distributed data mining technique that effectively handles the data generated by the nodes in the WSN and prolongs the lifespan of the network. Our work provides a novel cluster based scheme to mine the sensors data without moving it to cluster head (CH) or base station (BS) to achieve maximum performance in a WSN environment. The basic idea of the proposed work is that local computations are performed by utilizing the computing power at each sensor node and then the minimum higher level statistical summaries are exchanged, which decreases the energy dissipation in communication as the amount of the sensor data transferred is considerably reduced, and thereby the sensor network lifetime is maximized and also preserve the privacy of the sensor data.