Analysis, Comparison, and Optimization of Routing Protocols for Energy Harvesting Wireless Sensor Networks

Energy harvesting has been steadily gaining interest in the wireless sensor network community. Instead of minimizing the energy consumption and maximizing a network's operational time, the main challenge in energy harvesting sensor networks is to maximize the utility of the application subject to the harvested energy. One major challenge is to maximize the data delivery rates by exploiting the spatial variations of environmental energy. While there exists a multiplicity of energy-aware routing protocols for sensor networks without energy harvesting capabilities, only a small number of routing protocols have been published which explicitly account for energy harvesting. In this paper, we analyze and compare three state-of-the-art routing algorithms. While the original algorithms assume an idealized medium access control (MAC), a lossless wireless channel and global knowledge, we show that these assumptions lead to delusive results. We detail these findings by showing the influence of a low-power MAC protocol, a realistic wireless channel and the protocol overhead. Moreover, we show how to optimize the parameters of the MAC protocol for a given network configuration. By conducting various evaluations, we identify that our modified version of the R-MPRT algorithm outperforms the evaluated algorithms in scenarios where little energy is harvested from the environment.