A Knapsack-Based Message Scheduling and Drop Strategy for Delay-Tolerant Networks

Because of the dramatic changes in topology and frequently interrupted connections between nodes, messages in delay-tolerant networks are forwarded in the store-carry-forward approach. Routing methods in such an environment tend to increase the number of messages to improve the delivery ratio. However, excessive message copies lead to buffer overflows because of limited storage space. Therefore, an efficient message-scheduling and drop strategy is vital to maximizing network resources, especially when bandwidth is limited and message sizes differ. We developed a theoretical framework called the knapsack-based message scheduling and drop strategy in theory (KMSDT) based on epidemic message dissemination. To improve the delivery ratio, this strategy sorts message copies by utility per unit and, if buffer overflows occur, it decides which messages to drop based on the solution to the knapsack problem. Furthermore, we developed a practical framework called the knapsack-based message scheduling and drop strategy in practice (KMSDP). Rather than collecting global statistics as done in the KMSDT, KMSDP estimates all parameters by using locally collected statistics. Simulations based on synthetic trace are done in ONE. Results show that, without affecting the average delay or overhead ratio, KMSDP and KMSDT achieve a better delivery ratio than other congestion-control strategies.