Capacity and delay analysis in delay tolerant network with multiple communities

• This paper gets the maximal capacity of DTN with multiple communities.
• This paper proposes an algorithm to achieve the maximal capacity.
• This paper proposes a new algorithm to decrease the delay at the cost of capacity.

This paper studies the capacity and delay of delay tolerant network (DTN) with multiple communities. Due to the impact of social selfishness, nodes in DTN are more willing to forward messages coming from the same community. In such a scenario, this paper first mathematically defines the maximal capacity, and then proposes a single copy two-hop routing algorithm which can achieve the maximal capacity. Furthermore, it proposes a theoretical model to evaluate the delay of this algorithm. To decrease the delay, this paper modifies the above algorithm, and assumes that nodes just forward messages to nodes in the same community. The capacity and delay of this new algorithm are also obtained. Numerical results show that the selfish behavior decreases the capacity and increases the delay rapidly. However, the new algorithm can decrease the delay at the cost of decreasing the capacity. This result shows that if the sources just forward their messages to nodes in the same community, the delay may decrease.