Markov Lifetime Prediction Model for Wireless Sensor Network

Abstract

The Wireless Sensor Network (WSN) consists of sensor nodes that can communicate with each other via wireless transceivers to gather information from the environment. Sensor nodes are often placed in critical areas that are difficult to reach, sometimes in large numbers, and are unpredictable when the battery runs out on WSN. Thus, replacing the sensor node battery is difficult or even impossible. On the other hand, frequent battery replacement due to critical battery performance makes it uneconomical and unmeasurable. This paper proposes a lifetime battery prediction model for WSN. What is new in this study is that the battery lifetime is influenced by three factors, namely temperature, humidity, and data movement. Temperature and humidity significantly affect the battery lifetime, while data movement has little effect. Predictions of battery types one to four, with energy depletion (ED) consisting of Temperature, Humidity, and data movement. Each battery with these characteristics has a different lifetime; type one battery has a lifetime of 0.5 years, type two battery has a life of 0.16 years, type three has a life of 0.17 years, and type four has a lifetime of 0.4 years.