A analysis of machine learning in wireless sensor network

Abstract

Advances in hardware as well as wireless network tools have positioned us at the doorstep of a new-fangled era where undersized wire-less devices will endow with access to information every time, everyplace plus enthusiastically contribute in constructing smart atmos-phere. The sensors in WSN's assemble information regarding the substances they are exploited to sense. Nevertheless these sensors are restricted in their performance by restrictions of power plus bandwidth. Machine Learning methods can facilitate them in overcoming such restrictions. During the past decade, WSNs have seen progressively more rigorous implementation of highly developed machine learning algorithms for information handing out and improving network performance. Machine learning enthuse countless realistic clari-fications that make best use of resource exploitation along with make longer the existence of the network. In particular, WSN designers have effectively agree to machine learning paradigms to deal with widespread purposeful problems associated to localization, data ag-gregation, fault detection, Security, node clustering, prediction models and energy aware routing, etc.