The evaluation of communication networks and systems often mentions scalability as a key targeted property. Thus, a stringent definition of scalability and its quantification play a crucial role. In this chapter, the notion of scalability as well as the differentiation to related terms is revisited, before use case driven definitions of scalability are reviewed. The scalability index (SI) is then introduced and its application is demonstrated for the example of cloud gaming, for which two different strategies how to operate the streaming server are considered. Thereby, Quality of Experience (QoE) has manifested itself in research and industry, such that service providers can differentiate themselves by consistently delivering superior QoE compared to their competitors. The presented scalability framework allows comparing which system scales better in terms of QoE. To this end, fundamental relationships for quantifying QoE in systems are introduced. They allow measuring QoS in systems and deriving QoE metrics using appropriate mapping functions. For our use case, the average video bitrate experienced by a user is the QoS parameter, which is a main QoE influence factor. The fundamental QoE relationships allow mapping QoS to QoE metrics like Mean Opinion Scores (MOS), Good-or-Better (GoB) ratio, or Poor-or-Worse (PoW) ratio, which are discussed in the numerical results of the scalability analysis. Finally, recommendations and guidelines are provided on how to use SI in practice.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Scalability of Networks and Systems

  • Tobias Hoßfeld,
  • Poul E. Heegaard,
  • Wolfgang Kellerer,
  • Samuel Kounev

摘要

The evaluation of communication networks and systems often mentions scalability as a key targeted property. Thus, a stringent definition of scalability and its quantification play a crucial role. In this chapter, the notion of scalability as well as the differentiation to related terms is revisited, before use case driven definitions of scalability are reviewed. The scalability index (SI) is then introduced and its application is demonstrated for the example of cloud gaming, for which two different strategies how to operate the streaming server are considered. Thereby, Quality of Experience (QoE) has manifested itself in research and industry, such that service providers can differentiate themselves by consistently delivering superior QoE compared to their competitors. The presented scalability framework allows comparing which system scales better in terms of QoE. To this end, fundamental relationships for quantifying QoE in systems are introduced. They allow measuring QoS in systems and deriving QoE metrics using appropriate mapping functions. For our use case, the average video bitrate experienced by a user is the QoS parameter, which is a main QoE influence factor. The fundamental QoE relationships allow mapping QoS to QoE metrics like Mean Opinion Scores (MOS), Good-or-Better (GoB) ratio, or Poor-or-Worse (PoW) ratio, which are discussed in the numerical results of the scalability analysis. Finally, recommendations and guidelines are provided on how to use SI in practice.