Message Overhead Control Using P-Epidemic Routing Method in Resource-Constrained Heterogeneous DTN

Abstract

Delay Tolerant Networks (DTNs) are a kind of networks that are defined by their sporadic connectivity in their end to end path. Therefore, routing in these environments is based on the replication protocols such as epidemic routing. This protocol sends a copy of a message to other nodes that do not have it. This replication process demands the high amounts of network resources such as the node's buffer. Current researches studied the performance of such routing assuming that the nodes are homogeneous with the same transmission ranges and buffer sizes. In contrast, this work supposes that the nodes to have various buffer sizes and various transmission ranges. In DTN, the node's buffer plays a crucial role in delivering the messages. The nodes with limited buffer size cannot accept a new message and lead the network to loss those messages. As a matter of fact, the network acts as it is congested. This work proposes p-epidemic forwarding scheme to address this limitation. The number of redundant messages and accordingly, the buffer overflows are controlled by assigning the different transmission probabilities to the different nodes. The performance of our buffer control scheme by p-epidemic forwarding method is evaluated through simulations. Furthermore, the message delivery probability and the overhead ratio of our scheme is compared with epidemic routing.