Now a days Wireless Sensor Network is very hot topic for research due to its vast application in various area in our environment. It is Made up many independent motes distributed and interconnected in ad hoc manner for monitoring the environment continuously. The motes can be homogenous as well as heterogeneous according to the application or requirement. Today along with the many application of wireless sensor network there is also many obstacle of implementing it arising, which is very challenging in environment. The first and important challenge is how we can conserve energy in wireless sensor network. There is many algorithm and technic for conservation of energy and many of are going on to more optimize the motes in wireless sensor network. We have to need study about coverage in wireless network, Coverage means as how well and for how long sensor is able to monitor a field of interest. Several coverage practices have been researched depending on the area to be covered (area versus point coverage) and sensor deployment mechanism (discrete versus random). Link prediction algorithms often rely on network features and historical interaction data to make predictions. Common techniques include measuring network proximity, similarity metrics, and machine learning models In this paper we introduce the wireless sensor network in various perspective and its architecture and we will give model which consume less energy and excluding it to do unwanted work in particular domain, means we will discuss about converse factor so that the sensor can sense the medium accurately and appropriately.