低功耗物联网数据采集系统研究

Research on Low-power IoT Data Acquisition System

为了降低无线传感器网络的功耗,本文利用传感器采集数据通常具有较强的连续性和自相关性的特点,采用基于时域窗口的时间序列递进自回归结合粒子滤波的预测方法来准确预测数据的线性与非线性变化,减少传输数据量。同时,系统采用可配置的误差阈值,以控制数据收发端的数值误差,保证采集数据的准确性。使用LoRa无线通信协议进行园区车辆监测数据传输实验,结果表明,系统传输数据与真实数据基本一致,且具有较高的准确率,当数据传输减少50%时,均方误差可控制在0.45以下。本文方法有利于降低数据采集系统的整体功耗。

In order to reduce the power consumption of wireless sensor,a lightweight low-power IoT data transmission technique is studied to reduce the amount of transmitted data by using a window-based time series progressive auto-regressive prediction method and particle filtering method based on the feature that the data collected by sensors usually have strong continuity and autocorrelation.Configurable error thresholds are used to control the numerical errors at the transceiver side of the data to ensure the accuracy of the collected data.The vehicle environmental monitoring data transmission experiments using LoRa wireless communication protocols,the result show that the data transmission can be reduced by more than 50%while MSE can be controlled under 0.45.The method can reduce the overall power consumption of data acquisition system.