Elastic Optical Network Virtual Link Embedding Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current elastic optical networks face challenges in efficiently adapting to dynamic bandwidth demands due to limitations in flexible spectrum allocation and resource reallocation, leading to potential disruptions in traffic and increased operational costs.
Innovation Solution
A method and apparatus for dynamically adapting virtual links in optical networks by generating candidate embeddings based on network topology and current embeddings, determining total costs including disruption, transponder, and frequency slot allocation costs, and selecting the most optimal embedding to minimize disruption and resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed-grid spectrum allocation with 50 GHz or 100 GHz channels is used, then network infrastructure is simplified, but spectrum utilization efficiency deteriorates when bandwidth requirements are not aligned with fixed grid sizes
Solution Approach 1:
The patent segments the spectrum into fine-grained frequency slots of 12.5 GHz each, allowing flexible combination of multiple slots to match varying bandwidth requirements. This segmentation enables efficient spectrum utilization without requiring complex fixed-grid infrastructure, as slots can be dynamically allocated and released based on actual traffic demands.
Solution Approach 2:
The patent implements dynamic spectrum allocation where the bandwidth of virtual links can be adjusted in real-time by allocating or releasing frequency slots based on changing traffic demands. This dynamic approach allows the network to adapt to heterogeneous and variable traffic patterns, improving spectrum utilization efficiency while maintaining a relatively simple infrastructure.
2Quantity of substance
If multiple 50 GHz or 100 GHz channels are aggregated to meet larger bandwidth requirements, then bandwidth capacity is increased, but spectrum utilization efficiency deteriorates due to wasted capacity when only small bandwidth increases are needed
Solution Approach 1:
By dividing the spectrum into 12.5 GHz frequency slots, the patent enables granular bandwidth allocation. Instead of aggregating fixed 50 GHz or 100 GHz channels, the system can allocate exactly the number of slots needed (e.g., 2 slots for 25 GHz, 5 slots for 62.5 GHz), eliminating spectrum waste while providing flexible bandwidth capacity expansion.
Solution Approach 2:
The patent applies different allocation strategies to different parts of the spectrum based on local traffic requirements. Each frequency slot can be independently allocated to different virtual links, allowing precise matching of bandwidth capacity to actual needs in each network segment, thereby improving overall spectrum utilization efficiency.
3Adaptability or versatility
If advanced techniques such as push-pull defragmentation and hop retuning are used for bandwidth increase requests, then bandwidth flexibility is improved, but device complexity and operational difficulty increase due to developmental stage limitations
Solution Approach 1:
The patent uses fine-grained 12.5 GHz frequency slot segmentation as a foundational mechanism that enables bandwidth flexibility without requiring complex advanced techniques. By allocating individual slots rather than large fixed channels, the system achieves adaptability through simple slot allocation and release operations, avoiding the need for developmental techniques like push-pull defragmentation.
Solution Approach 2:
The patent introduces a network controller as an intermediary that manages frequency slot allocation centrally. This controller simplifies the implementation of bandwidth flexibility by handling slot allocation, release, and reconfiguration automatically, avoiding the need for complex distributed coordination mechanisms like hop retuning while maintaining adaptability.
4Adaptability or versatility
If dynamic resource allocation in EONs is implemented to address diverse traffic demands, then adaptability is improved, but disruption to existing traffic increases during reconfiguration
Solution Approach 1:
The patent implements a two-stage adaptation process where candidate embeddings are generated and evaluated before actual reconfiguration. By preliminarily assessing multiple candidate solutions and selecting the one with minimum disruption, the system can adapt to traffic demands while minimizing impact on existing flows. This preliminary evaluation ensures that reconfiguration actions are optimized to protect traffic continuity.
Solution Approach 2:
The patent uses disruption cost as a feedback metric in the embedding selection process. By calculating and comparing disruption costs of different candidate embeddings, the system receives feedback on potential traffic impact and selects configurations that minimize disruption. This feedback mechanism enables adaptive resource allocation while maintaining traffic reliability.
Data Source
AI summary
The disclosed apparatuses and methods are directed to embedding of virtual links in an optical network. The method comprises: receiving an adaptation request for a virtual link within a virtual network embedded on an optical substrate network; generating a plurality of candidate embeddings based on a topology of the substrate network and a current embedding of the virtual link, each candidate embedding satisfying the adaptation request; determining a total cost of each candidate embedding based on a disruption cost of the candidate embedding; and selecting, as a new embedding, a candidate embedding from the plurality of candidate embeddings in accordance with the determined total cost of the selected candidate embedding.