A stepwise optimization algorithm of clustered streaming media servers

  • Authors:

  • Yunpeng Chai;Zhihui Du;Yinong Chen

  • Affiliations:

  • Tsinghua National Laboratory for Information Science and Technology, Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China;Tsinghua National Laboratory for Information Science and Technology, Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China;Computer Science and Engineering Department, Arizona State University, Tempe AZ 85287, USA

  • Venue:
  • Journal of Systems and Software
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The optimization of Clustered Streaming Media Servers (CSMS), which aims at using as few hardware resources and as cost-effective as possible, while providing satisfactory performance and QoS, has a great impact on the practicability and efficiency of CSMS. Based on the analysis and formulization of critical performance factors of CSMS and the relationship among the performance, QoS, and the costs in CSMS, a stepwise optimization algorithm is developed to solve the optimization problem efficiently. The algorithm is based on an approach that models the optimization problem into a directed acyclic graph and then addresses the complex optimization problem step by step. The algorithm applies a divide and conquer model that not only reduces the complexity of the optimization problem, but also accelerates the optimization process. Progressive information is collected in the process and used in solving the problem. Furthermore, a simulation system of CSMS is necessary for the optimization algorithm to generate the accurate information produced in the entire streaming service process. Thus, we designed and implemented such a simulation system based on the theoretical performance model of CSMS and the parameters measured in practical CSMS testbed. Finally, a case study of the optimization problem is given to demonstrate the process of the algorithm, and an appropriate plan for designing practical CSMS system is illustrated.