Search

By collecting and analyzing various data from things connected to the Internet, IoT can create new valuable information. Then, such obtained information is transmitted to an actuator, which can perform suitable operation to the situation and condition on the spot. However, there are only a few case studies focusing on how information obtained by data collection and analysis is given to each actuator. Therefore, in this research, we aim at achieving a network that distributes information according to requirements of actuators. To achieve this purpose, we considered applying Directed Diffusion, which is a Publish / Subscribe communication model. However, when the conventional Directed Diffusion is applied to the assumed environment of this research, there are issues that many interest and exploratory packets are generated, and that the data packet transfer route is frequently disconnected due to the movement of IoT devices. Therefore, we propose a method to connect the new data packet transfer route to the existing data packet transfer route used by another actuator. In this method, when there is a reinforced gradient corresponding to interest, transfer of interest packet is stopped, and exploratory packet is transferred from that node. In this case, since transfer of these packets is performed between the actuator and the intermediate node, the number of these transfers is suppressed. Further, it is possible to re-establish the data packet transfer route in a short time. [ABSTRACT FROM AUTHOR]

In conventional network science, we have considered the network as a single graph so that analyzed its characteristics. As a result, "scale-free" and "small-world" were discovered. However, it can be thought that the network exists not as a single, but as it is related to each other. The relationship between the networks makes a phenomenon that can't be explained by looking at a single network. A representative example of this phenomenon is the massive blackout that occurred in Italy in 2003. This is the incident that was caused by the mutual dependence between the electricity grid and the control network forming the electric power network. When constructing a network, it is common to make it resistant to failures. Therefore, we want to eliminate the possibility that the fault tolerance of the system will decrease due to the relationship between the networks. Then, in this research, we propose the model of occurrence and propagation of faults on interdependent networks, which is one example of networks that exist in relation to each other. Also, as an example of an interdependent network, we refer to software defined network (SDN). [ABSTRACT FROM AUTHOR]