۱- Introduction

۲- Related work

۳- System model

۴- Joint data gathering and energy harvesting scheduling

۵- Near-optimal buffer-battery-aware scheduling

۶- Performance evaluation

۷- Conclusion and future work

References

Abstract

In this paper, we study the joint data gathering and energy harvesting (JoDGE) problem in rechargeable wireless sensor networks (RWSNs) with a mobile sink. In RWSNs, the sensor nodes are equipped with RF circuit to harvest energy from a mobile sink that moves along a pre-defined path, and at the same time, transmit gathered sensor data to the sink. Given that the consumed and harvested energy at a sensor node is proportional and inversely proportional to the square of transmission distance, a far-relay approach is proposed to select the sensor nodes closer to the path to assist the data transmission of the farther sensor nodes. Under the far-relay approach, we formulate a network utility maximization problem (NUM), and propose an optimal scheduling scheme (Opt-JoDGE), which jointly considers the power allocation, relay selection and time slot scheduling policies. By employing the Lyapunov drift theory, a near optimal buffer-battery-aware adaptive scheduling (NO-BBA) scheme is further proposed, in which the run-time status of the data buffer and battery are utilized. Extensive simulation experiments validate the feasibility and performance of JoDGE and NO-BBA. The results show that the performance of NO-BBA is close to that of Opt-JoDGE, especially when a certain delay is tolerable.

Introduction

One of the challenges in wireless sensor networks (WSNs) is how to gather data from sensors through a resource-constrained wireless network [9]. In WSNs, the sensor nodes periodically sample the physical entities under monitoring, and then transmit the gathered sensor measurements to a sink, which is connected to the rest of the system for data processing and decision making. In order to improve the sustainability, recently, various energy harvesting technologies have been employed in WSNs [2, 14, 17, 19, 22]. This kind of WSNs is referred to as rechargeable WSNs (RWSNs). Extensive research efforts have been devoted on effective data gathering in RWSNs. For example, [14] proposed an energy-efficient cooperative data collection scheme for clustered RWSNs. In [17], an optimal scheduling algorithm was proposed to minimize data packet loss in RWSNs, where the sink is assumed to be a fixed station. In RWSNs with static sinks, the transmissions of sensor data to the sinks may pass through one or multiple relay nodes. Thus the sensor nodes geographically closer to the sink usually have to forward more sensor data. Therefore, they are more likely to become the bottleneck of the network due to heavy relay workload. By contrast, data gathering in RWSNs with mobile sink(s) has been shown to be a promising approach to jointly deal with unbalanced traffic distribution and prolong the network lifetime. The mobile sink is assumed to travel along a pre-defined or online-learned path, and the network throughput maximization (NTM) problems were investigated, e.g., through routing and time-slot scheduling [30, 38], joint speed and power control [23], and mobility planning [6, 28, 39, 40]. These works assumed that the sensor nodes harvest energy from either unstable environment sources, i.e., solar and wind, or energy based on magnetic resonance coupling with small charging distance.