Summary: Wireless Sensor Networks consist of low power nodes with limited reception distance. In these networks, Internet of Things based various protocols are used to meet quality criteria such as stability and reliability that industrial applications must provide. Of these protocols, the one responsible for routing operations is the RPL protocol. Since the nodes are located without knowing the network structure, they have to choose a parent node when they join the network. Due to communication problems in the network, nodes have to change their preferred parents over time. Bad parent choice or frequent parent changes negatively affect the efficiency of the network. Nodes in the network cannot provide the same quality of communication throughout a period hour, day, month, etc. . In this study, a new learning based objective function named LBHOF is proposed for the RPL protocol running on the IETF 6TiSCH protocol stack. For a realistic approach, some nodes are set to different transmit power in certain period slices. After the tests, a significant improvement was achieved in the number of parent switch, packet delivery ratio and the number of packets dropped from the buffer memory. It has been observed that iterative running networks with the proposed objective function behave more stable.
Key Words: Wireless Sensor Networks, Internet Of Things, IETF, 6TiSCH, RPL,
Learning Based, Objective Function. |