Probability based Admission Control for Guaranteeing Latency Constraints of Multiplexed VBR Traffics

As a next generation service, immersive media is the most attractive one, which is indispensable for supporting human tele-presence and virtual lab. The immersive media service requires high standards on interactive response delay and synchronization among media for comfortable communication in users' perspective. To measure the customer's experiences from the network provider side, we propose a new measure of quality of experience (QoE), the distortion ratio of service datum (DRS), by calculating how much original statistical property of the service datum is distorted at the receiver end. DRS can make possible compare the QoEs among the heterogeneous services by providing the deterministic value. We also suggest a connection admission control algorithm for supporting the stringent QoE in the virtually synchronized packet network. With the statistical on-off flow description, the loads on a link after aggregating traffic flows can be estimated in terms of probability. According to the level of QoE, admission control algorithm guarantees the deterministic latency in probabilistic manner, while maximizing the usage of network resources.