Passive Diagnosis for Wireless Sensor Networks

Network diagnosis, an essential research topic for traditional networking systems, has not received much attention for wireless sensor networks (WSNs). Existing sensor debugging tools like sympathy or EmStar rely heavily on an add-in protocol that generates and reports a large amount of status information from individual sensor nodes, introducing network overhead to the resource constrained and usually traffic-sensitive sensor network. We report our initial attempt at providing a lightweight network diagnosis mechanism for sensor networks. We further propose PAD, a probabilistic diagnosis approach for inferring the root causes of abnormal phenomena. PAD employs a packet marking scheme for efficiently constructing and dynamically maintaining the inference model. Our approach does not incur additional traffic overhead for collecting desired information. Instead, we introduce a probabilistic inference model that encodes internal dependencies among different network elements for online diagnosis of an operational sensor network system. Such a model is capable of additively reasoning root causes based on passively observed symptoms. We implement the PAD prototype in our sea monitoring sensor network test-bed. We also examine the efficiency and scalability of this design through extensive trace-driven simulations.