论文部分内容阅读
接纳控制是视频服务中保证客户服务质量的重要机制。在介绍一种新的分布式视频点播系统的基础上,将客户分为不同的优先级类,每类客户采用基于阈值的接纳控制方法,分析了各个客户类的稳态阻塞概率,并利用Kaufman/Roberts递归算法来降低计算复杂度。建立了系统收益模型,从系统收益最大化的角度,给出了确定各个客户类最佳阈值的方法。仿真结果与理论分析完全吻合,验证了分析模型的有效性和正确性,同时也证明该算法能够保证高优先级的客户优先享用服务器资源,并且就系统收益而言,该算法优于完全共享的接纳策略。
Admission control is an important mechanism in video services to ensure customer service quality. Based on the introduction of a new distributed video on demand (VOD) system, the clients are divided into different priority classes. Each type of customers adopts the threshold-based admission control method to analyze the steady-state blocking probability of each client class. Kaufman / Roberts recursive algorithm to reduce the computational complexity. The system revenue model is established. From the perspective of maximizing system benefits, a method of determining the optimal threshold for each customer class is given. The simulation results are in good agreement with the theoretical analysis, which verifies the validity and correctness of the analysis model. At the same time, it also proves that the algorithm can guarantee the high-priority customers to enjoy the server resources preferentially. And in terms of system benefits, the algorithm is superior to the completely shared Accept the strategy.