Article Preview
Top1. Introduction
Efficient wireless energy transmission technology in wireless rechargeable sensor networks (WRSNs) has developed rapidly for many years (Clerckx et al., 2019; Mudulia, L., Mishrab, D. P., & Jana, K., 2018; Liu, X., Guo, Y., Li, W., Hua, M., & Ding, E., 2018; Yao, K. H., Jiang, J. R., Tsai, C. H., & Wu, Z. S., 2017), which addresses the problems of charging, lifetime extension, and energy renewal between different devices. Recently, current research on WRSNs has focused mainly on the design and scheduling of wireless charging vehicles (WCVs) (Xu, C. J., Cheng, R. H., & Wu, T. K., 2018; Zhong et al., 2017; Lin et al., 2018).
Using a WCV to replenish energy for WRSNs requires addressing the problem of the energy limitation (Xie et al., 2015). In most current charging solutions, the WCV with wireless chargers must wait on site until the sensor is fully charged, which will result in long-term waiting and will increase the risk of sensor failure due to power exhaustion. Therefore, a partial charging mode was suggested so that the WCV can move to the next charging location without completing the current sensor's charging (Xu, W., Liang, W., Jia, X., & Xu, Z., 2016). This partial charging mode can also extend the WCV's running time as a result of less energy consumption on driving the WCV’s own motion. As an improvement, Cheng, R.-H. (2019) proposed another cooperative charging scheme with multiple WCVs to extend the whole lifetime of WRSNs, but it will lead to an increase in the costs of devices. However, the method using one or more WCVs still suffers from unbearably long waiting times of charging.
To further extend the lifetime of sensors, Xu et al. (2018) designed a novel type of WCV with a separable charger array, which can carry multiple chargers to replenish energy to adjacent sensors at the same time. As there is no need to stay on site, the WCV can continue to perform assigned charging tasks. This new concept of a separable charger array can greatly improve the rescue efficiency for sensors and can successfully extend the lifetime of entire WRSNs. Even so, the number of chargers that can be used in the array is limited, and the chargers need to be recollected when their charging services are finished.
Xu et al. (2018) proposed a backward recollection algorithm (BRA) to fulfill the charger array again for the next charging mission. According to the BRA, during its back way toward the base station (BS), the WCV will recollect chargers near the WCV’s moving line. However, according to the BRA, as the backward charger recollection method is a straightforward strategy, the effectiveness of recollection is low. There is an upper limit on the number of chargers that can be recollected. The recollection rate of chargers is lower than 30%.
However, consider that the WCV can pick up the chargers to be recollected, not only on the back way but also on its charging tour toward sensors. In this work, we propose a new recollection algorithm called the opportunistic charger recollection algorithm (OCRA). Unlike the BRA, which recollects chargers only on the back way of the WCV’s charging mission, in the OCRA, the WCV will perform recollection actions in the whole charging route to recollect chargers as many as possible.
The OCRA algorithm mainly includes four steps: (1) Segmenting the WCV’s charging route to restrict the scheduling computation range in the length direction. (2) Calculating the range of recollection in the width direction according to the lifetime of sensors, the speed of the WCV and other parameters. (3) Optimizing the order of chargers to be recollected in each segment of the WCV’s charging route. (4) Combining the results of all segments to merge into a whole scheduling route. Through the above four steps, the overall rate of recollected chargers is between 40% - 60%, which has greatly outperformed the previous achievements.
The rest of this paper is organized as follows. Section 2 is the review of previous work. Section 3 presents the proposed new recollection algorithm in detail, followed by a comparison and evaluation of simulation results in Section 4. Finally, Section 5 concludes this paper with further work.