Scholarlink

<p>This paper proposes a crosstalk-aware inter-core (XT) circuit reallocation algorithm for spatial division multiplexed elastic optical networks (SDM-EON). Unlike previous studies that utilize reallocation primarily for spectral defragmentation, this work focuses on circuit reallocation to mitigate XT, thereby reducing or preventing network blocking. The algorithm is triggered whenever a request is blocked, classifying it as a reactive approach. The push-pull and fast-switching techniques are employed for data traffic migration, ensuring seamless transition without service interruption. Furthermore, the proposed method is evaluated against other algorithms designed to mitigate inter-core crosstalk, considering the NSFNET, EON, and JPN network topologies. In terms of bandwidth blocking probability, the results demonstrate a reduction of at least 65%, with a maximum of 0.25% of active circuits reallocated per process.</p>

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

Crosstalk-aware circuit reallocation to reduce blocking in spatial division multiplexed elastic optical networks

  • 期刊论文

  • 发表日期：

    2025年4月22日

查看全文
  • Selles G. F. C. Araújo,
  • André C. B. Soares

摘要

This paper proposes a crosstalk-aware inter-core (XT) circuit reallocation algorithm for spatial division multiplexed elastic optical networks (SDM-EON). Unlike previous studies that utilize reallocation primarily for spectral defragmentation, this work focuses on circuit reallocation to mitigate XT, thereby reducing or preventing network blocking. The algorithm is triggered whenever a request is blocked, classifying it as a reactive approach. The push-pull and fast-switching techniques are employed for data traffic migration, ensuring seamless transition without service interruption. Furthermore, the proposed method is evaluated against other algorithms designed to mitigate inter-core crosstalk, considering the NSFNET, EON, and JPN network topologies. In terms of bandwidth blocking probability, the results demonstrate a reduction of at least 65%, with a maximum of 0.25% of active circuits reallocated per process.