Adaptive Energy Efficient RPL Avoiding Unstable Nodes for Smart Factory

In smart factory, many IoT devices detect events and process them. Most of these IoT devices need to operate wirelessly, so they use batteries and are configured to use wireless networks. However, these wireless factory networks are prone to be interfered by many environmental factors. Therefore, smart factories require an energy-efficient network of wireless IoT devices even in such harsh environments. RPL is a distance vector IPv6 routing protocol designed for Low-Power and Lossy Networks such as factory networks. However, RPL has a problem of high energy consumption due to frequent parent changes in a lossy network. In this paper, we propose ADP-RPL, which is an improved protocol of RPL to solve the aforementioned problem. The proposed ADP-RPL deals with the problem that RPL selects unstable nodes with weak link quality as the preferred parent just considering the current link quality. In order to avoid these unstable nodes, ADP-RPL reflects the deterioration of current link quality from the previous link quality and the network stability of potential parents. In this way, the child node maintains the routing stably by avoiding unstable nodes when selecting the parent node. In addition, ADP-RPL adaptively reflects the threshold of MRHOF. In the existing RPL, child nodes are affected by more frequent parent changes than their parents using the same threshold regardless of the value of rank. ADP-RPL limits the parent change using a larger threshold value for the neighbor with a higher rank. This adaptive threshold reduces energy consumption due to frequent parent changes. In Cooja simulator, we implement ADP-RPL on Contiki OS and show that it outperforms the existing RPL in the viewpoint of energy consumption.