AOF-RPL: Towards an MCDM Based Adaptive Objective Function in RPL

In recent years, many research works have focused on RPL, especially in relation to the Internet of Things. This makes RPL an ideal data collection protocol for lossy environments where nodes have low energy levels and limited capabilities. In practice, RPL builds a tree-like topology, often represented by a destination-oriented directed acyclic graph that aims to transmit data to an LLN border router. Many multimetric extensions of RPL are based on a set of predefined metrics and an objective function; each node involved in the topology construction process goes through a parent selection by considering static weights for the set of metrics. The topology is periodically rebuilt to maintain the network while considering the same tools. To provide more flexibility, we present in this paper AOF-RPL, an extension of RPL which is designed to adjust the relevance of each metric based on the network's global condition and QoS requirements of the running applications. The problem is formulated using an MCDM method called ISOCOV. Simulations performed on COOJA show that our approach improves globally the performances compared to similar protocols while enhancing the network lifetime.