Soft fault diagnosis in wireless sensor networks using PSO based classification

In this work, a real time soft fault diagnosis model is proposed for wireless sensor networks (WSNs) using particle swarm optimization (PSO) based classification approach. The proposed model follows in three phases such as initialization, fault identification, and fault classification phase to diagnose the composite faults (combination of soft permanent, intermittent, and transient fault) in the sensor network. The faulty nodes are identified in the network based on Analysis of variance (ANOVA) method. The feed forward neural network (FFNN) technique with PSO learning method is used for classification of the faulty nodes. We evaluate our model by carrying out the testbed experiment in an indoor laboratory environment.