论文部分内容阅读
该文提出了两种不同的策略 ,来处理一个基于层次型服务器网络的大规模 TV- Anytim e系统中的动态影像映射问题 .基于预测的阶段性重映射策略是每隔一个时间周期决定服务器网络中的影像对象映射 ,以处理系统较长时间内的变化 .自适应的重分配策略则处理系统较短时间内的变化 ,该重分配策略基于不同影像对象的收益 /开销 ,其目标是尽大化服务器系统产生的收益 .模拟实验表明 ,这些影像对象管理策略提高了系统的性能和自适应性
This paper proposes two different strategies to deal with the dynamic image mapping problem in a large-scale TV-Anytim e system based on a hierarchical server network.Predictive phased re-mapping strategy is to determine the server network every second period of time In order to deal with changes in the system over a long period of time.Adaptive redistribution strategy to deal with changes in the system within a short period of time, the redistribution strategy based on different image objects of the benefits / overhead, the goal is to maximize The revenue generated by the server system simulation experiments show that these image object management strategies to improve system performance and adaptability