| Many wireless environments face long delays and intermittent connectivity anomalies. A promising solution is delay-tolerant networks, but these networks have inherent issues with routing. Popular mobile ad hoc routing protocols such as AODV and OSPF rely on the existence of an end-to-end path. These protocols first establish a path before traffic is sent. A delay-tolerant network, however, is characterized by intermittent connectivity and consequently for these types of networks, traditional routing protocols fail. This thesis presents an adaptive threshold routing algorithm with the intent of improving upon the traditional threshold method [9]. An adaptive algorithm is an ideal approach for dynamic, challenged, network environments. It takes into account changes in the network structure and adjusts the algorithm parameters to optimize performance in terms of reducing network load, lowering the delivery rate of packets, and maximizing packet throughput. In order to analyze the performance of the adaptive threshold algorithm, both the threshold method and the adaptive threshold method are implemented in a discrete event simulator called OMNET++. Simulations are run to analyze the performance of both of these algorithms in varying network topologies and the results are compared. Experimental results show that as the network topology changes, the adaptive threshold method not only maintains a high throughput while lowering the network load, it also outperforms the threshold method by reducing the average packet delivery rate. |
