Data Fusion Method Exploiting Data Correlation for Physical Wireless Parameter Conversion Sensor Networks
Kohei Natsume, Mai Ohta, Osamu Takyu, Takeo Fujii · 2020
Physical wireless parameter conversion sensor net-works (Phy-C-SN) were proposed to gather information at high time efficiency. In this method, a fusion center can simultaneously receive multiple signals from sensor nodes (SNs). It is because each SN transmits a subcarrier of orthogonal frequency division multiplexing (OFDM) which corresponds to its sensing information. It is important for wireless sensor networks to decrease the power consumption of SNs in order to prolong the network lifetime. In the proposed method, SNs are divided into clusters based on the similarity of their sensing values and only an SN of each cluster transmits its data. The SNs in the same cluster transmit in turns so that their power consumptions are fair. In a fusion center, the missing data is interpolated with the data from SNs belong to the same cluster. The computer simulation results show that the number of transmission is decreased while the error is suppressed in this scheme. If the allowable error is set 0.2, 0.5 °C, the total number of transmissions within the network is decreased by 20 % and 31 % in the proposed method compared with that of the conventional method and 98.1 % and 97.8 % of the transmissions are kept less than or equal to the acceptable error, respectively.