Reliability analysis for various communication schemes in wireless CORBA
Reliability analysis for various communication schemes in wireless CORBA
复制标题
DOI：
10.1109/tr.2005.847268
复制
发表时间：
2005-05
影响因子：
5.9
通讯作者：
Xinyu Chen;Michael R. Lyu
中科院分区：
文献类型：
--
作者：
Xinyu Chen;Michael R. Lyu
为了设计更可靠的网络，我们将传统的可靠性分析从有线网络扩展到组件不完善的无线网络。无线网络系统，如无线CORBA，继承了独特的切换特性，导致了不同组件和链路的不同通信结构。因此，传统的双端可靠性定义已不再适用。我们提出了一个新的术语，端到端期望瞬时可靠性，将这些不同的通信结构整合到一个度量中，该度量不仅包括故障参数，还包括服务参数。然而，它仍然是一个随时间单调递减的函数。对不同无线通信方案的端到端期望瞬时可靠性及其相应的MTTF进行了定量评估。为了观察整体可靠性提高的增益，还对不完美部件的可靠性重要性进行了评估。结果表明，不同部件的失效参数对MTTF和可靠性重要性的影响是不同的。随着系统预期工作时间的不同，可靠性改进的重点应从一个部件转移到另一个部件，以获得最大的可靠性增益。此外，在通信状态期间参与的组件的数量比系统状态的数量更为关键。为简单起见，我们假设有线和无线通信链路是完美的，在可靠性分析中忽略它们。如果将这两者结合到所提出的端到端期望瞬时可靠性中，可以对无线网络系统进行更详细、完整的可靠性评估。我们的定量测量作为一个例子，假设故障率和维修率是恒定的；然而，在实践中，故障和服务过程可能遵循其他分布。毕竟，我们的调查提供了一个初步但全面的方法来衡量无线网络的可靠性。虽然我们的分析是在无线CORBA平台上进行的，但它很容易扩展到一般的无线网络系统。
For the purpose of designing more reliable networks, we extend the traditional reliability analysis from wired networks to wireless networks with imperfect components. Wireless network systems, such as wireless CORBA, inherit the unique handoff characteristic which leads to different communication structures with various components & links. Therefore, the traditional definition of two-terminal reliability is not applicable any more. We propose a new term, end-to-end expected instantaneous reliability, to integrate those different communication structures into one metric, which includes not only failure parameters but also service parameters. Nevertheless, it is still a monotonously decreasing function of time. The end-to-end expected instantaneous reliability, and its corresponding MTTF, are evaluated quantitatively in different wireless communication schemes. To observe the gain in overall reliability improvement, the reliability importance of imperfect components are also evaluated. The results show that the failure parameters of different components take different effects on MTTF & reliability importance. With different expected working time of a system, the focus of reliability improvement should change from one component to another in order to receive the highest reliability gain. Furthermore, the number of engaged components during a communication state is more critical than the number of system states. For simplicity, we assume that the wired & wireless communication links are perfect, and omit them in the reliability analysis. If these two are engaged into the proposed end-to-end expected instantaneous reliability, it can give a more detailed & complete reliability assessment of a wireless network system. Our quantitative measurements are conducted as an example with the assumption that the failure & service rate are constant; however, in practice, failure & service processes may follow other distributions. After all, our investigation provides an initial yet overall approach to measure the reliability of wireless networks. Although our analysis is conducted on wireless CORBA platforms, it is easily extensible to generic wireless network systems.