Performance Evaluation of Highly Efficient Information Collection Methods by Trend Analysis of Sensor Information Using Pre-learning

Abstract

LPWA (Low Power Wide Area), a low-power, long-distance communication method, has been attracting attention. Level Index Modulation (PLIM) has a problem of missing information due to packet collision when multiple sensors select the same index. In this study, we established an optimal design based on a mathematical model that minimizes the probability of packet collision by using a trend analysis of prior sensor information in the index design, which is the correspondence between packet transmission time, selected channel, and transmitted information. In this paper, we proceeded with the characterization by computer simulation and showed that the proposed design method can suppress the packet collision probability better than the conventional method.