Rethink data forwarding in mobile social networks using movement history information

This paper studies data forwarding by using the node’s spatial information in mobile social networks (MSNs). Specifically, we partition the 2D space into several grids, and periodically record the nodes’ staying within each grid to extract their movement history summaries. Then, nodes’ movement history summaries are used to compare their forwarding abilities in the single-copy scenario. In the multiple-copy scenario, we first address the dependent data forwarding path problem, i.e., different copies will reach the same relay with good forwarding ability, and thus the advantage of multiple-copy cannot be fully utilized. To avoid this, we jointly consider the nodes’ forwarding abilities and their movement trajectories to perform copy distribution. Therefore, the potential overlap of multiple copies is minimized. In addition, we propose an extended scheme, which periodically records the nodes’ transaction in grids. It improves the performance at the cost of more computation and storage consumption. Through extensive trace-driven experiments, proposed algorithms achieve good performance in different scenarios.