Data Dissemination in Delay Tolerant Networks

Delay Tolerant Networks (DTNs) are characterized by large delays, frequent disruptions and lack of contemporaneous paths between nodes. Moreover, nodes may have limited computational power, storage, and battery capacity. Despite these challenges, DTNs have found applications in areas such as animal tracking, surveillance and facilitating education in remote areas. In these applications, a critical problem is routing or data dissemination. Indeed, any routing protocols must aim to transmit data efficiently and maximize data delivery. The challenge, however, is that nodes are only connected intermittently, and do not have a contemporaneous path. Moreover, any established path may not remain valid after a transmission, and nodes may experience large delays before encountering one another. As a result, existing routing protocols cannot be retrofitted to work in DTNs. To date, a primary solution for routing in DTNs is epidemic-based routing protocols because of their simplicity, low delays and little to no reliance on specific nodes. A key observation of past research on epidemic routing protocols is that they are not evaluated on a unified framework. Specifically, no work has compared the performance of epidemic routing protocols using both the Random Way Point (RWP) model and trace files. Henceforth, this thesis has conducted a comprehensive study that employs both RWP and trace files. In particular, this thesis is the first to compare the performance of epidemic-based routing protocols using a custom-built simulator that moves nodes according to a trace-file and the RWP model. Moreover, it compares these protocols using