摘要

Rapid developments in network technologies and mobile devices improve our lives, but the resultant fast-growing traffic leads to a lack of cellular network resources and breaks the balance between profit and costs for network providers. Data ofloading can relieve the loads on network infrastructure and shift data to a complementary wireless technology, thereby increasing maximum throughput and extending network coverage. In this paper, we propose a terminal-to-terminal-based ofloading method called Virtual Bank with movement prediction (VBMP), which uses an opportunistic network to achieve ofloading. We establish a virtual trading system to manage information transmission. Meanwhile, node movement prediction and node subset selection via a submodular function are used to select the relay nodes and to discover the dynamic topology of mobile nodes. Theoretical analysis and simulation results demonstrate that VBMP effectively improves data ofloading.