Proper Buffer Size and Time to Live on the Scheduled Train System using Delay-Tolerant Networks

In general, the simulation results of DTNs were delivery probability graphs. In general, the graph was obtained from variations in the number of messages and variations in buffer sizes. Some graphs were obtained by varying bandwidth sizes, variations in number of nodes, and variations in the number of hops. In this research, the delivery probability graphs were obtained from the variation in message sizes, variation in buffer sizes, and variation of TTL values. The test used the ONE simulator and protocol routing PRoPHET and PRoPHETv2. From these variations, the graphs of the message size of 1 Kilo Byte were similar and were not depend on the variation in buffer size but the graphs of the message size 10 Kilo Bytes and 100 Kilo Bytes varied. There was a graph that showed a maximum delivery probability value at TTL 480 minutes and TTL 780 minutes for all buffer sizes and was not increase with a larger TTL. Another graph showed the maximum delivery probability value in TTL 960 minutes and in TTL smaller than 960 minutes. Some graphs showed the maximum delivery probability value in certain TTL values that depends on the buffer size.