Abstract
Wireless rechargeable sensor networks (WRSN) can prolong the lifetime of wireless sensor networks (WSN). In WRSN, mobile charger (MC) is used to charge low-power sensor node (SN). How to plan an efficient charging path for MC is a NP-hard problem. The task is to plan a path for MC that can not only charge the SN requesting charging in time, but also reduce the driving distance, so as to reduce the energy loss of MC. In this paper, policy gradient path planning (PGPP) algorithm based on deep reinforcement learning is proposed to solve this problem. In PGPP, the SNs in WRSN are treated as the environment, and the base station and MC are seen as the sensors and actuators of the agent. A reward function which can measure the energy efficiency of MC is designed, and the policy gradient algorithm is used to learn the strategy. Simulation experiments show that compared with other MC charging path planning algorithms, PGPP algorithm has certain advantages in the speed of path planning and the energy efficiency of WRSN.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call