Client-class based admission control for distributed Video-on-Demand systems

Admission control is one important Quality of Service (QoS) mechanism for maintaining performance of streaming media service. The focus of this paper is to analyze the client-class based admission control scheme for distributed Video-on-Demand (VoD) systems. In order to assure specific QoS, clients are classified into K priority classes based on their contributions in terms of profits, and blocking probabilities of different client classes are calculated through the recursive algorithm. Also the paper presents an adaptive threshold policy based on client request time to better improve the resource usage. Numerical analysis shows that the client with higher priority does have more chances of accessing the video server resources, thus maximizing the system´s profits.