Smart blacklisting: A solution to increase performance of blacklisting in LOADng for low density networks

Abstract

G3-PLC is emerging as an attractive communication standard for smart metering applications. This standard includes different mechanisms and protocols allowing to deal with the poor transmission performance of the power line channel. In particular, all nodes of the network can act as relay. In 2014, G3-PLC standard, the ITU-T G.9903 recommendation, proposed the LOADng protocol to perform the routing. Blacklisting is a mechanism included in this protocol that deals with bad quality links by ignoring them. However, the blacklisting mechanism suffers from a lack of performance when the density of nodes is too low. Indeed, some bad quality links are the only ones possible to reach some farther nodes so that by blacklisting them, these farther nodes cannot be reached anymore. This is why a solution had to be developed in order to solve this problem. Our solution is the smart blacklisting mechanism that highly reduces the probability of blacklisting a critical link. In this work, we highlight the problem of blacklisting in low densities and compare the performance of classical blacklisting and smart blacklisting. For that purpose, we developed our own simulator in the Contiki/Cooja platform in which we have implemented the LOADng protocol as well as the different features of G3-PLC networks. Based on our tool, we report the results we obtained for different node densities and for different link qualities with the classical blacklisting mechanism, without it and with the smart blacklisting mechanism.