An Enhanced Differential Private Protection Method Based on Adaptive Iterative Wiener Filtering in Discrete Time Series

Abstract

Although many proposed researches on differential privacy protection in correlation time series have made great progress, there are still some problems. Because different methods are based on different models and rules. There is no uniform attack model, their privacy protection intensity cannot be compared and measured horizontally. This paper designs an attack model for the differential privacy in correlation time series based on adaptive iterative wiener filtering. Experimental results show that the attack model is effective and provides an uniform measurement for the privacy protection with different methods.