The Blocking Option in Routing Protocols

Routing protocols are designed under the assumption that each node in a network should be able to reach (i. e. send or forward packets to) every other node in the network. Unfortunately, adopting this assumption in a routing protocol does allow adversary nodes to launch spam or DoS attacks against the other nodes in the network. In this paper, we introduce the "blocking option" in routing protocols; this option allows a node u to block a specified set of nodes {v,..,w} and prevent each of them from reaching node u. It turns out that if node u blocks a large number of nodes, then u may end up blocking other nodes as well. We refer to these unintentionally blocked nodes as blind_to u nodes. Clearly, a node u cannot communicate with its blind nodes in a regular manner. Thus, we extend the routing protocol to allow each node u to communicate with its blind nodes via some special node, called the joint node. To perform its intended function, the joint node needs to be neither blocked by any node nor blind to any node in the network. We give an algorithm for identifying the node that is best suited to be the joint node in a network. Finally, we show, through extensive simulation, that the average number of blind nodes is close to zero when the average number of blocked nodes is small (< 20) and that the probability of a joint node being blind is quite small, on the order of 10^-3. The path length of using a joint node for communication between a node u and any one of its blind nodes v is around 1.5 the shortest path between u and v.