摘要:中国电信云化IP城域网自2020年大规模商用以来,宽带业务网元从传统的多业务边缘设备(MSE)架构逐步演进为集中池化部署的云化虚拟宽带远程接入服务器(vBRAS)架构。然而,vBRAS集中池化部署的可靠性研究缺乏理论验证。聚焦于量化城域网部署模式的核心机制,深入剖析业务故障类型,建立了vBRAS及MSE部署模式的可靠性概率模型,同时辅以严格的数理公式推导。创新性地提出以故障概率与受灾用户数的概率分布作为可靠性量化指标,并结合数理模型与当前现网部署实例进行分析。研究结果表明,云化vBRAS部署模式在可靠性与成本效能方面均显著优于传统MSE部署模式,为后续城域网建设提供了理论支持。
关键词:可靠性;云化IP;城域网;池化;vBRAS;概率论
Abstract: Since the large-scale commercial deployment of China Telecom's cloudified IP metropolitan area network in 2020, the broadband service network elements have gradually evolved from the traditional Multi-Service Edge (MSE) architecture to a centralized and pooled cloud-based virtual Broadband Remote Access Server (vBRAS) architecture. However, research on the reliability of centralized and pooled vBRAS deployment lacks theoretical validation. Focusing on quantifying the core mechanisms of metropolitan area network deployment models, this study conducts an in-depth analysis of service failure types and establishes reliability probability models for both vBRAS and MSE deployment modes, supported by rigorous mathematical derivations. Innovatively, it proposes using the probability distribution of failure rates and the number of affected users as quantitative reliability indicators, combining mathematical models with real-world deployment examples for analysis. The research results demonstrate that the cloudified vBRAS deployment mode significantly outperforms the traditional MSE mode in terms of both reliability and cost efficiency, providing theoretical support for future metropolitan area network construction.
Keywords: reliability; cloud IP; metropolitan area network; pooling; vBRAS; probability