I. Introduction
Nowadays, Wireless Sensor Network as an important technology for IoT, is more and more widely used in various research fields. Most of them are adopted to track and monitor different kind of things such as environment, building facilities, plants and animals. Specific applications include building condition monitoring [1], urban facility monitoring [2], smart grid monitoring [3], eco-environment monitoring [4], precision agriculture [5], and human health monitoring [6], etc. Moreover, these applications have a common requirement that the sensor network needs to work continuously for a long time. However, the battery capacity is small due to the limited sensor node size, which needs to be charged above the threshold. It is very difficult for thousands of wireless sensor nodes to get their battery recharged. So the charging capability of wireless sensor network limits the reliability, lifetime and application area. In order to solve the energy limitation, three solutions can be adopted, namely energy-saving method [7], energy harvesting method [8] and wireless charging method [9], respectively. The energy-saving method sacrifices certain network performance, such as increasing network delay, reducing data reliability, etc. It only saves limited energy. The energy harvesting method requires bulky collection equipment and cannot ensure the charging quality. Environmental energy harvesting is limited to surroundings environment and sometimes reduces the energy deficiency. The wireless power transmission technology can be divided into three categories: inductive coupled, electromagnetic radiation and magnetic resonant coupled [10]. Inductive coupling-based WPT is mainly used for near-field transmission and the transmission distance is usually limited to millimeters or centimeters [11]. The main advantage of electromagnetic radiation approach is its long transmission distance but there are radiation concerns. Therefore, it is mainly used in special fields such as solar satellites.