Research Area:  Internet of Things
RPL, the standard IPv6 routing protocol for low-power and lossy networks in the emerging Internet of things (IoT), is designed mainly for efficient many-to-one data collection scenarios where the majority of the traffic flows from embedded devices to a gateway. Although RPL does support root-to-node downwards routing and peer-to-peer (P2P) communication, its P2P performance is inefficient and unsatisfactory due to excessively high churn and bottlenecks in the P2P path. However, P2P routing is important for machine-to-machine communication where nodes in IoT applications communicate with one another and control devices beyond simply collecting data. In this work, we propose a neighbor-graph-based RPL (NG-RPL) that significantly improves P2P routing performance. By including additional routing information when a packet passes through the root node for the first time in a P2P communication, NG-RPL finds efficient P2P routes opportunistically, when available, without significant overhead. We implement NG-RPL in Contiki-NG and evaluate its performance through extensive Cooja simulations. Results show that NG-RPL reduces routing churn, which improves packet reception ratio, round-trip time, and energy usage of P2P communication compared to standard RPL.
Keywords:  
Author(s) Name:  Yongjun Kim; Jeongyeup Paek Chung-Ang University, Seoul, South Korea
Journal name:  IEEE Access
Conferrence name:  
Publisher name:  IEEE
DOI:  10.1109/ACCESS.2020.3028771
Volume Information:  Volume: 8
Paper Link:   https://ieeexplore.ieee.org/document/9214466