A Review of ML Based Fault Detection Algorithms in WSNs

Wireless sensor network (WSN) is precisely outlined as a group of exclusively dedicated spatially distributed sensors for recording and processing environmental data like temperature, humidity, wind velocity, air density etc. WSN is a propitious technology because of its cost effectiveness, facile deploybility and flexible size. But, because of several reasons sometimes the WSN changes dynamically and it demands various advanced algorithms and at times, redesigning of the network architecture. ML techniques prove to be helpful in coping up with these disruptive changes. Machine learning is a self-learning approach that enables computing machines to learn from their experiences and respond without the requirement of any human trainer or re-programming [1]. In this paper, we have compared several ML algorithms that work well for fault detection in WSNs.