Wireless Sensor-Networks Conditions Monitoring and Fault Diagnosis Using Neighborhood Hidden Conditional Random Field

This paper formulates wireless sensor networks (WSNs) fault diagnosis problem as a pattern-classification problem and introduces a newly developed algorithm, neighborhood hidden conditional random field (NHCRF), for determining hidden states between sensors. The health conditions of WSN are determined by using the NHCRF model to estimate the posterior probability of different faulty scenarios. The NHCRF model can improve the WSN fault diagnosis, because it has relaxed the independence assumption of the hidden Markov model. To enhance the robustness and antinoise ability of the NHCRF, the concept of nearest neighbors is used when estimating dependencies. In this paper, a 200-sensor-node WSN is used to show that the proposed NHCRF method can deliver excellent and effective results for WSN-health diagnosis. Our study also presents thorough results on different types of WSN traffic, the free traffic, light traffic, and heavy traffic. Comparative results indicate that our method can deliver superior classification performance compared with other methods.