紫金山实验室：2023未来网络白皮书-光电融合服务定制广域网_63页_6mb

Summary of White Paper Analysis

Background and Challenges

Networks face new demands from scenarios like industrial internet, computing power internet, and consumer applications, needing guaranteed Quality of Service (QoS). Electrophotonic integration is vital for evolving WANs to handle multi-dimensional QoS requirements, moving from bandwidth-focused to service-customization models.

Traditional Technologies

Traditional WAN technologies, such as optical transport and data networking, have limitations in capacity, flexibility, and QoS assurance due to independent development of optical and electronic domains. Examples include IPoWDM, IPoOTN, FlexE, DIP, and CSQF, which address some issues but still lack full integration and deterministic capabilities.

Industry Solutions

Existing industry approaches, such as Cisco's RON, Juniper's CORA, Huawei's NCE, and ZTE's SDN-based schemes, demonstrate partial electro-optic integration but primarily focus on bandwidth or coarse-grained services. These solutions simplify architectures but fail to provide fine-grained, multi-dimensional QoS commitments.

CustomWAN Innovation

CustomWAN introduces a three-layer architecture integrating packet, TDM, and optical layers for electro-optic fusion. It enables multi-dimensional QoS customization through unified scheduling, resource pooling, and deterministic forwarding, supporting Network as a Service (NaaS). This approach offers expressive differentiation and flexible network control for emerging applications.

Economic Models

Traditional pricing models (e.g., flat-rate or usage-based) are inadequate for QoS-service differentiation. A new model based on multi-dimensional QoS metrics is proposed, allowing users to pay for assured performance like latency, jitter, and bandwidth, incentivizing efficient network resource use.

Summary of Strengths and Limitations

Strengths: Achieves high resource efficiency, large-scale transmission, and end-to-end QoS commitments, aligning with strategies like "East Data West Calculation" and digital economy goals.
Limitations: Increased complexity in networking and control planes, requiring advanced synchronization and coordinated resource management. Future enhancements include improving security analytics, reducing technical barriers, and optimizing economic models for broad adoption.