Reliability and energy-efficiency communication are two main design constraints for wireless sensor networks (WSNs). The BATS research project targets at tracking bats living in their natural habitat and to detect social interactions of the animals. Tiny mobile nodes are mounted on the bats and ground nodes are distributed over their habitat and together span a WSN. The physical constitution of bats highly limits the size and weight of the mobile nodes and consequently the battery capacity. Moreover, the communication between mobile and ground nodes is presumably highly spontaneous and unreliable, because of natural barriers and the mobility of flying bats. Enabling a reliable communication, despite of unreliable channels, is an energy-intensive task.

Commonly used approaches like automatic repeat request (ARQ) suffer from increased latency and the requirement of a feedback channel. An alternative to overcome reliability problems without these disadvantages is forward error correction (FEC). Up to now a lot of research has been performed in terms of communication latency and performance, but only little work has been done in examining the energy consumption of FEC techniques.

In this work we evaluated FEC techniques with respect to the energy consumption and the amount of restored data. Two different error models are considered: one similar to the one of the BATS scenario and a second one based on a bit-error probability model. Our evaluation results show, that in both scenarios high data restoration rates are achieved with less energy consumption compared to common approaches, hence the lifetime of the WSN can be increased.

BIO

Benedict Herzog is a member of the System Software Group at the Friedrich-Alexander University of Erlangen-Nürnberg (FAU) and currently works towards his Master's degree in Computer Science. His main research topics are distributed systems with sensor nodes and corresponding operating systems aspects.