Research on Distributed Differential Desensitization Algorithm for Wireless Sensor Network

Abstract

Aiming at the problems of large noise variance and poor validity of desensitized multi-dimensional data using the differential desensitization algorithm in wireless sensor network, a personalized distributed differential desensitization algorithm based on random response mechanism and segmentation mechanism is proposed. The algorithm realizes the switching between the random response mechanism and the segmentation mechanism by setting the value of the desensitization budget, and selects different disturbance domains through random sampling values. In order to further protect the data, the desensitization process is designed in the data collection stage, and the sink node directly desensitizes the data collected from the sensor, and then transmits it in the network. Combining different transformation strategies and perturbation methods for sensor data of wireless sensor network, and selection of desensitization timing, the algorithm improves the desensitization effect of multi-dimensional sensor data in wireless sensor network, and improves data validity through random sampling mechanism. Through theoretical analysis and simulation, the noise variance obtained by the personalized distributed differential desensitization algorithm based on random response mechanism and segmentation mechanism for multi-dimensional sensor data is better than the existing desensitization algorithms, and has better data validity.