An Efficient Architecture for Imputing Distributed Data Sets of IoT Networks

Abstract

In the era of the Internet of Things (IoT), spatially distributed IoT devices collect and store data in a distributed fashion for computational efficiency. However, in IoT networks, due to the fragile device, harsh deployment environment, and unreliable transmission, the possibility of missing data is increasing, which may significantly affect subsequent data processing. Traditional approaches to impute missing data in IoT distributed datasets bring huge communication overheads. In this paper, we develop an efficient architecture for distributed IoT data imputation based on a designed multi-discriminator conditional generative adversarial network. The architecture intelligently learns the characteristics of the distributed datasets to accurately impute missing values. Our experiments are performed using three datasets under two different data missing mechanisms. The experimental results demonstrate that using three datasets, the proposed imputation technique can drastically reduce the imputation error by up to 88.66%, 94.27%, and 95.53% at the premise of low transmission cost, respectively, compared to five state-of-the-art methods.