Towards a Model Checker for NesC and Wireless Sensor Networks

Wireless sensor networks (WSNs) are expected to run unattendedly for critical tasks. To guarantee the correctness of WSNs is important, but highly nontrivial due to the distributed nature. In this work, we present an automatic approach to directly verify WSNs built with TinyOS applications implemented in the NesC language. To achieve this target, we firstly define a set of formal operational semantics for most of the NesC language structures for the first time. This allows us to capture the behaviors of sensors by labelled transition systems (LTSs), which are the underlying semantic models of NesC programs. Secondly, WSNs are modeled as the composition of sensors with a network topology. Verifications of individual sensors and the whole WSN become possible by exploring the corresponding LTSs using model checking. With substantial engineering efforts, we implemented this approach in the tool NesC@PAT to support verifications of deadlock-freeness, state reachability and temporal properties for WSNs. NesC@PAT has been applied to analyze and verify WSNs, with unknown bugs being detected. To the best of our knowledge, NesC@PAT is the first model checker which takes NesC language as the modeling language and completely preserves the interrupt-driven feature of the TinyOS execution model.