An Energy-efficient Data Collection Scheme by Mobile Element based on Markov Decision Process for Wireless Sensor Networks

Wireless sensor networks allow efficient data collection and transmission in IoT environments. Since wireless sensor networks consist of a few to several sensor nodes which are spatially distributed, the data collection in the networks is a difficult task. Due to multi-hope connectivity and energy constraint, mobile elements are sent to the network to collect data from these sensor nodes directly by one hope communication. The mobile elements need an efficient moving strategy to minimize the data-gathering latency and energy consumption while maximizing the rate of gathering data. The goal is to estimate an effectual movement policy of the mobile elements that improve the rewards and data collection rate in non-stationary environments. In the proposed scheme, a movements policy for the mobile elements is formulated through the Markov decision process. The computer simulation shows that the proposed scheme significantly improves the data gathering rate and avoids the buffer overflow condition of the sensors to reduce the data loss and energy consumption of sensor nodes and mobile elements.