AMNA: Probe Agent Based Inter-Process Dependency Model for Wireless Sensor Network’s Fault DiAgnosis

Process failures in communication protocol stack of wireless sensor networks (WSNs) are caused by underlying errors and faults. This work presents AMNA: probe Agent based inter-process dependency Model for wireless sensor Network’s fault diAgnosis. The probe agent performs online process execution tracing to analyze the effect of errors, omissions and channel misbehavior on process execution at node level. An inter-process dependency model is proposed that is implemented through a decision tree for fault diagnosis. AMNA has been implemented and evaluated on Castalia simulator.