Congestion management in opportunistically connected networks using repeated games

In opportunistically connected networks, custody transfer provides a certain degree of reliability as a custodian node cannot discard a message unless its life time expires or the custody is transferred to another node with a commitment. This creates a challenging decision making problem at a node in determining whether to accept a custody transfer: on one hand, it is beneficial to accept a large number of messages as it can potentially advance more messages toward their ultimate destinations and network utilization can be maximized; on the other hand, if a receiving node over-commits itself by accepting too many messages, it may find itself setting aside an excessive amount of storage and thereby preventing itself from receiving further potentially important messages. In this paper, we employ repeated single-decision games to model this dynamic behavior and propose congestion management strategies. The repeated game based approach is particularly suitable for opportunistically connected networks, where every node cannot be sure that it is going to communicate with different opponents next round and there is no knowledge when the dynamic behavior is going to stop. The proposed congestion control strategy is completely distributed in nature where only the local information such as the occupied storage space of a node is required. Our simulation results also show that the proposed control strategy is effective in avoiding congestion and balancing network load among the nodes.