Improving communication quality of internet of things for pelagic fishing vessel by blended fog computing and cloud computing

Abstract: The communication bandwidth from pelagic fishing vessel to shore was unable to cope with the increasing needs of smart devices and sensors on vessel. Fog computing and cloud computing were combined to solve this problem effectively in operating system of pelagic fishing vessel. Cloud computing could be used as the basic framework of system. Fog computing could be used as model in the local area network and sensor network on vessel. This mixed computing architectures could effectively play the advantages of each other and have a complementary effect. In research, cloud computing was taken as basic framework of operating system of pelagic fishing vessel, and service-oriented architecture service component was used to provide a reliable and safe data storage center, which could reduce the equipment requirements for client and share data and applications between different devices. There were two sources for service components. One part came from the encapsulation of original system function and was released through the network via the standards such as simple object access protocol, web services description language, universal description discovery and integration, et al. . It could effectively protect the investment. Other parts were new service components facing new requirements, which were packaged by extensible web service and supported open, dynamic interoperability model. The standard cloud computing framework increased the demand for network bandwidth. For the data volume of firmware update and message sent by sensors occupied larger proportion of communications from vessel to shore, this paper also studied equipment firmware distribution mechanism and shipboard sensor network computation model based on fog computing in order to reduce ship shore data traffic. With the aid of computing and storage capacity of shipboard equipment, the firmware distribution mechanism changed the traditional way that each device got update files from the cloud directly, pushed and updated the firmware between smart devices and sensors in local area network of vessel. This paper also studied shipboard sensor network computation model based on fog computing. For shore-based command center just was focused on changes or change tendency of message sent by sensors, the system could reduce the amount of data transmission by send filtered date only. Smart devices with computing and storage capacity could be used to select and calculate message data before they were sent to cloud by sensors in vessel, which reduced the amount of communication with high monitoring accuracy. Practice has proved that cloud-fog mixed computing architectures could not only ensure interactive integration of information from vessel to shore, but also significantly reduce the requirement for data communication bandwidth and the network traffic. During the practice period, from March 15 to May 31 in 2014, the data volume of firmware update file of smart devices and sensors was 250.905 MB, the actual firmware update communication flow was 37.175 MB. Firmware update data decreased to 14.81% of the traditional architecture. At the same time, the data volume of cumulative message produced by 27 temperature sensors was 571.63 MB, and the actual data volume of message sent from vessel to shore was 27.16 MB after selection and calculation, so message data volume reduced to 4.75%. In this period, about 7500 were saved. According to this calculation, each boat can save communication cost by 36 000 each year and the costs of communication were markedly reduced. The empirical research obtained obvious effect on a pelagic fishing vessel named Kaifu.