Waqaas Munawar, Olaf Landsiedel, Muhammad H Alizai, and Klaus Wehrle (2009)
Remote Incremental Adaptation of Sensor Network Applications
In: Proceedings of the 8th GI/ITG KuVS Fachgespräch "Wireless Sensor Networks", Hamburg, Germany.
Wireless Sensor Networks (WSNs) are deployed for long periods of time, during which a need often arises to dynamically reprogram or retask them. An array of solutions has been proposed to this effect, ranging from full image replacement to virtual machines. However, the capabilities of TinyOS – the current state of the art in sensor node operating systems – are still limited to full image replacement. TinyOS based applications have a modular architecture but during compilation this modularity is lost resulting in a statically linked system image.
In this work we extend TinyOS to allow dynamic exchange of components in WSN applications by conserving their modularity during the compilation process. This generates the possibility of incremental adaptation of sensor nodes’ behavior through partial code replacement. The designed system does not require any alterations in the existing user interfaces, remaining transparent to the user. The evaluation shows that our approach imposes almost no performance overhead for loaded application while keeping a smaller memory footprint than other comparable solutions.


