Lost in Space Or Positioning in Sensor Networks

We discuss the success chances of a real-world implementation of a positioning system for a wireless sensor network. While much re- search has been done in the area of node localization, the method of choice to verify the results has been in theory or by simula- tion. To realize the visions of future sensor networks, we take very basic sensor nodes (prototyping the smart dust idea) and implement a series of experiments to determine their suitability for use in positioning algorithms. Specifically, we look at the ac- curacy and effectiveness of direct distance measurements based on different signal strengths. We found that even though links are stable and symmetric over time, positioning in the real world is yet in its infant stage and that current theoretical models of sensor networks do not apply well in unspecialized hardware.