Improving information quality of sensory data through asynchronous sampling

In this paper, asynchronous sampling is proposed as a novel approach to improve the information quality of sensory data through shifting the sampling moments of sensors from each other. The exponential correlation model and the entropy model for the sensory data are introduced to quantify their information quality. An asynchronous sampling strategy, EASS, is presented accordingly to assign equal time shifts to sensors, which in turn reduces data correlation and thus improves information quality in terms of increased entropy of sensory data. A lower bound for EASS is derived to evaluate its effectiveness. Simulation results based on both synthetic data and experimental data are satisfactory.