Energy saving is one of the most important issues for Internet of Things (IoT). An intuitive way to save energy of IoT devices is to reduce the reporting frequency to the IoT server. However, to do so, the time-variant values are distorted, which may be influential to the measured results. In this paper, we take PM2.5 application as an example to discuss the relation between energy efficiency and data accuracy. Through analyzing PM2.5 concentration collected via LoRa at National Chiao Tung University (NCTU) from 2016 to present, two reporting mechanisms based on timer and threshold, respectively, are proposed. The experimental results demonstrate that the threshold-based reporting outperforms the timer-based reporting by more than 37% in energy saving when the accuracies of these two reporting mechanisms are the same.

节能是物联网（IoT）最重要的问题之一。节省物联网设备能源的一种直观方法是减少向物联网服务器报告的频率。但是，这样做会使时变值失真，这可能会对测量结果产生影响。本文以PM2.5应用为例，探讨能效与数据精度之间的关系。通过分析2016年至今通过交通大学LoRa采集的PM2.5浓度，提出了两种分别基于定时器和阈值的报告机制。实验结果表明，当这两种报告机制的精度相同时，基于阈值的报告在节能方面比基于计时器的报告要高出37％以上。