Scholarlink

<p>Optical networks are designed to handle wireless traffic by interlinking heterogeneous uplink and downlink communication networks. This work introduces an Augmented Wavelength Division Allocation Method (AWDAM) using a perceptron learning model. The proposed allocation method probabilistically estimates the in-link traffic flow over the different peak intervals. The perceptron learning estimates the allocation and acceptance capacity of the interlinking devices to disperse the peak traffic. The learning augments verify the interlinking device’s capacity to dissolve the peaks through mesh-interconnected routing. More specifically, the wavelength augmentation and division for joint traffic management using multiple interlinking points is preceded based on the blocking probability. This blocking probability reduction for heterogeneous outflows is ensured by both the perceptron layer decisions during the peak interval. Therefore, the traffic grooming is augmented and divided according to the layer’s decision that jointly improves wavelength allocations. The traffic grooming is pursued aided by dual operations of routing and allocation, reducing the blocking rate by 10.79% for the maximum device load. This method is efficient in improving resource allocation and network throughput by 13.31% and 11.28% for the same device loads.</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

Augmented wavelength division allocation method for peak traffic grooming in optical networks using 2-layer perceptron learning

  • 期刊论文

  • 发表日期：

    2026年1月27日

查看全文
  • J. Kumarnath,
  • J. Vetrimanikumar

摘要

Optical networks are designed to handle wireless traffic by interlinking heterogeneous uplink and downlink communication networks. This work introduces an Augmented Wavelength Division Allocation Method (AWDAM) using a perceptron learning model. The proposed allocation method probabilistically estimates the in-link traffic flow over the different peak intervals. The perceptron learning estimates the allocation and acceptance capacity of the interlinking devices to disperse the peak traffic. The learning augments verify the interlinking device’s capacity to dissolve the peaks through mesh-interconnected routing. More specifically, the wavelength augmentation and division for joint traffic management using multiple interlinking points is preceded based on the blocking probability. This blocking probability reduction for heterogeneous outflows is ensured by both the perceptron layer decisions during the peak interval. Therefore, the traffic grooming is augmented and divided according to the layer’s decision that jointly improves wavelength allocations. The traffic grooming is pursued aided by dual operations of routing and allocation, reducing the blocking rate by 10.79% for the maximum device load. This method is efficient in improving resource allocation and network throughput by 13.31% and 11.28% for the same device loads.