A Modified RPL for Wireless Sensor Networks with Bayesian Inference Mobility Prediction

Wireless Sensor Networks (WSNs) become one of the most common technologies that can be deployed in various domains. However, the networks suffer from many problems caused by the limited sensor resources and the harsh environments where these networks are deployed. Many algorithms have been proposed to manage data routing while respecting the specificity of such networks. One of the known proposed routing protocols is RPL (IPv6 Routing Protocol for Low power and Lossy networks). Nevertheless this protocol is designed with consideration of the limited sensor energy, it cannot be integrated in may WSN applications. In fact, RPL, as it has been proposed, assume that the sensor nodes are static and don't manage any type of mobility. In this paper, we propose a new approach, called BMP-RPL (Bayesian model Mobility Prediction RPL for wireless sensor networks), that aims to adapt native RPL to nodes' mobility scenarios. This approach is based on nodes' identification and velocity prediction as well as the estimation of the link duration. Thus, we introduce a new metric which constructs routes according to the node status and within information lost. The performance of our approach is proved through simulation.