Perceiving who and when to leverage data delivery for maritime networks: An optimal stopping view

The advances in the integration of wireless communication and sensor technologies have stimulated an innovative paradigm named Crowd Sensing Networks, which caters to the exponential growth of service demands on the sea and drives the development of potential maritime wideband networks. This paper investigates the issue of sensed traffic data scheduling between vessels, combining Time Division Long Term Evolution (TD-LTE) and delay-tolerant networks (DTNs) on the sea. Specially, we propose a unique network topology which combines maritime crowd sensing network and delay tolerant networks, i.e., a store-carry-and-forward routing topology is explored to address the intermittent network connectivity in maritime context. Notably, the alternative eco-friendly green energy in maritime environment will make the scheduling issue more challenging. To the best of our knowledge, this is the first work to do such investigation with the goal of minimizing the costs associated with end-to-end delay of data delivery and energy consumption of DTN throw box. On this basis, we design the scheme of data cooperation transmission between vessels that the hosting vessel decides which DTN throw box to store the data, and when a vessel arrives, the DTN throw box determines whether to stop, i.e., let the arriving vessel carry the data, or skip it and continue to wait for other vessels. A Two-step Time and Energy Oriented Optimal-stopping (TTEOO) algorithm leveraging backward induction method is proposed, based on the optimal stopping rules. Simulation results are presented to show the effectiveness of the proposed method, in terms of consumption cost and data delivery ratio.