Undervolting in WSNs - A long-term study of IdealVolting by using the outdoor testbed 'PotatoNet'
The intention of this evaluation was to show whether undervolting is usable for WSNs or not. For this purpose we deployed PotatoNet and performed a long-term study of about 4 months. The WSN was continuously toggled between an undervolted state and a state where all nodes were powered by the recommended voltage level. Within each state, every node broadcasted a message on 16 different tx-levels. Every event (IdealVolting output, transmitted messages, received messages) were logged by the central box for post-processing.
Experiment
The PotatoNet was deployed on a trial field from the VSD research station for potato crops. Hence, our evaluation was performed in the light of a real-world Smart Farming application with farming and harvesting activities.
Field deployment: 07/17 - 08/21
Deployment on the field with potato crops between the nodes. The deployment took place during a normal season. Hence, also farming activities or sprinkling affected the WSN. The VSD research station noted activities like sprinkling or heavy rain. A report is available on request.
Field deployment: 08/25 - 09/11
A haulm chopper removed the potato plants on the surface. To avoid a damage on the PotatoNet testbed, we temporary removed the field nodes. After this procedure we relaunch the evaluation with the same topology as before.
Edge deployment: 09/11 - 11/18
Due to final harvesting activities we had to move the nodes from the trial field to the edge of the field. The evaluation ended on november 18. We 'harvested' arround 7.5GByte of data we like to share with the community.
Data
All data that have been collected during this evaluation are available for download here.
More Pictures
Some interesting pictures from construction, deployment and operation of Potatonet can be found in our gallery.
