Streaming
STREAMING FLOW ANALYSIS
What kind of research?
The videos you usually watch on You tube are called streaming communications. Unlike data communications such as e-mail, these communications require quality control techniques to ensure communication because data is constantly being received. In our laboratory, we are theoretically analyzing network control schemes that enable users to watch streaming communications without complaints.
In addition, by incorporating the user behavior of users of these communications (e.g., the user's willingness to wait a little longer before connecting in case of congestion or emergency, or the user's willingness to connect at a lower bit rate) into the theory, we can design a system that encourages cooperative user behavior. We are considering the possibility of doing so.
Resource allocation for Streaming communication
We propose an acceptance/rejection control method that maximizes user satisfaction by considering the requested bit rate of streaming communication with availability in the network. Assuming a probability distribution of when users request videos and when they finish watching, we theoretically analyze which and how much video should be accommodated in the network in an environment where various video bit rates are mixed.
MPEG DASH bit rate selection method
We propose a method to determine the bit rate for MPEG DASH, a protocol used in various video services, while maximizing the user utility, Quality of Experience (QoE). The buffer state is analyzed using queueing theory or game theory.