Data traffic interruption detection and mitigation framework for RPL-based WSNs

Wireless Sensor Networks (WSN) monitor various aspects of the environment in which they are deployed. Sensor nodes collect and send environment data by neighboring nodes until it reaches the gateway node (root node). WSNs are susceptible to several attacks that totally or partially interrupt the data flow, such as sinkhole, blackhole, gray hole/selective forwarding. We here present an application layer framework for RPL-based WSNs that detects and mitigates the attacks that interrupt data traffic while informing the root node (and the network maintainers) of the suspect nodes by alternative routes, thus preventing interception by attackers. The attacker nodes are then relegated to leaf-nodes by the local reaction of legitimate nodes, as they are prevented from acting as routers, effectively mitigating attacks. By concentrating monitoring functions on the sink, network managers can identify compromised nodes, which can lead to their physical removal. The experiments show that our framework kept the package loss rate at about 2% or lower in most of those tested. Besides managing to maintain a low control message overhead (i.e., 4.67% in a no attack scenario and 9.74% under attack) it was able to provide the network managers with information about the location of the attackers.