No abstract

The critical applications of wireless sensor networks, the increased data faults and their impact on decision making reveal the importance of adopting online techniques for data fault detection and diagnosis. Keeping in mind the hardware limitations of sensors, this work focuses on complementary signal processing techniques (temporal, spatial correlation and self organizing map) in order to cover several types of data faults, reduce the misdetection rate and also isolate faults when possible by specifying the defaulting sensors. The methods applied to a real database show that 31.6% of data are faulty by applying SOM3D in conjunction with the spatial correlation. The combination of the above technique in addition to the temporal correlation reduces the misdetection by increasing the detection percentage by 17.6%. SOM3D model also helped identifying the least trustful sensors among the network sensors, this can be helpful when reconciling errors.

Sensor faults are the rule and not the exception in every WSN deployment. Sensors themselves may get stuck at a particular value or get partially disconnected and report noisy measurements. Sensor nodes may reboot unexpectedly or stop transmitting data. Software running on the sensor nodes may have bugs and may cause data loss. In this paper, we present an efficient approach for online data fault detection and its application to a case study from a real world dataset. Our approach exposes four types of data faults as they occur by locally applying five simple heuristic rules. By locally applying these rules, node will not need to exchange messages with neighboring ones and consequently to consume energy. Simulation results showed that around 19% of the total collected readings from a real world dataset were faulty.

Safety critical applications, such as explosion prediction, require continuous and reliable operation of Wireless sensor networks (WSNs). However, validating that a WSN system will function correctly at run time is a hard problem. This is due to the numerous faults one may encounter in the resource-constrained nature of sensor platforms together with the unreliability of the wireless links networks. A holistic fault tolerance approach that addresses all fault issues does not exist. Existing fault tolerance work most likely misses some potential causes of system failures. In this paper, we propose an integrated fault tolerance framework (IFTF) that reduces the false negative by combining a network diagnosis service (component/element level monitoring) with an application testing service (system level monitoring). Thanks to these two complementary services, the maintenance operations will be more efficient leading to a more dependable WSN. From the design view, IFTF offers to the application many tunable parameters that make it suitable for various application needs. Simulation results show that the IFTF reduces the false negative rate of application level failures to 60% with an increase of 4% in power consumption (communication overhead) compared to using solely network diagnosis solutions.

No abstract