Abstract
Shipyards are potential environments for using IoT services, sensor networks, and delay tolerant networks. Simulations of those services and networks strongly rely on human mobility models. Results obtained with an unrealistic model may not reflect the true performance of applications, protocols, and algorithms in a shipyard. A lot of synthetic models for human movements have been studied but most of them are generic and focus on the daily movements of humans on city scales. Nevertheless, workers in shipyards have unique movement characteristics such as movement speed, pause time, and attractions places. For instance, workers usually move to some places, where they work, and rarely move to other places in the factory. Movement characteristics of workers not only depend on workers but also on tasks, which they do. For instance, workers, who paint ships, have similar movement speed and pause time. Hence, in this paper, human movements in shipyards are studied. We propose a new human mobility model called the human mobility mode in shipyards (MIS). In MIS, workers are classified into multiple types. Movement characteristics of a worker are similar to other workers in the same type. Based on the visiting probability, workers have some places, where they frequently visits, and some places, where they rarely visit. We analyze real mobility traces and studie to achieve human movement characteristics from real traces. The results show that MIS provides a well-match to the movement characteristic from real traces.