Performance analysis of multichannel random access for spatial correlated source in narrowband Internet of things system

Abstract

Recently, the global warming phenomena become more severe and the scale of climate change increase, which have a major impact on human life. However, existing environmental observations have limitations in early warning and forecasting. Because, conventional observation station covers large area and are located in a certain area (e.g., urban area, building rooftop). Therefore, it is inevitable that countermeasure is delayed in an area far from the observation stations. To solve this problem, wireless sensor networks have been used for acquisition of precise climate observations. Although many sensors can be used for early warning and prediction, many collisions which disturbs the observation transmissions occur in the random access procedure.

In this thesis, we studied NB-IoT random access for climate observation transmissions that makes measurement precision with spatial correlation in wireless sensor networks. First, we analyzed the performance of NB-IoT in terms of throughput and delay time. In addition, a Arduino based IoT platform was constructed to acquire correlated data. Then, based on the acquired data, spatial correlation distance was calculated using the T-test method which used for multichannel random access. The system considering the spatial correlation distance can reduce the redundancy of the information, so that the influence due to the collision can be reduced. Finally, we derived the recovery performance of multichannel random access considering spatial correlation distance and mobility of moving objects. From the analysis and simulation results, we calculated the number of moving objects in the Gwangju Institute of Science and Technology (GIST) campus to achieve the desired recovery performance. In addition, we compared the efficiency of the proposed system with the fixed system from an economic point of view.