Multivariate Data Acquisition Strategy by Employing Signal Processing on Graph in Edge Computing Based Industrial Internet of Things

Multivariate data acquisition strategy is a crucial issue in edge computing based Industrial Internet of Things (EC-IIoT), which should be carefully designed to improve energy efficiency and prolong network lifetime of EC-IIoT. Sampling on graph provides a promising opportunity to solve resource limited sensor selection problem via forming the appropriate sampling set and reconstructing original graph signal. In this paper, we translate the multivariate data acquisition strategy to the graph sampling set forming problem, which could exploit the adjacent correlation of the vertices in the spatial and variable domain. The simulation result demonstrates the effectiveness of the proposed data acquisition strategy.