Mid-America Transportation Center
Date of this Version
2010
Document Type
Article
Citation
Report # MATC-UNL: 226 Final Report 25-1121-0001-226
Abstract
The freight railroad industry in North America is exerting efforts to leverage Wireless Sensor Networks to monitor systems and components on railcars. This allows fault detection and accident prevention even while a train is moving. Railcars, constructed mostly of ferrous materials such as steel, are expected to severely impact signal propagation. To better understand this effect we first evaluated the signal characteristics when sensor nodes are placed in various locations around railcars. We used EM-Field modeling and evaluation techniques to obtain these results and found that node placement selection is critical for the node’s communication distance. As a second research area we therefore aimed at understanding the protocol requirements and limitations of current WSN technologies. Based on the results of our study we found ZigBee to be inadequate for freight WSNs and developed a solution that remedies the problems we observed. Our evaluation of this new multi-tier approach shows a significant performance and network lifetime gain, making freight train wireless sensor networks feasible.
Comments
Copyright 2010 Mid-America Transportation Center