Optical Burst Switching Convergence with Traffic Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optical burst switching (OBS) convergence methods based on adaptive length thresholds have limited adaptability and precision, leading to suboptimal dynamic adjustment of burst length thresholds, which affects traffic processing efficiency and service differentiation.
Innovation Solution
A service convergence method and system that uses a traffic prediction mechanism to calculate and adjust the convergence threshold based on predicted burst demand, allowing for adaptive precision and flexibility in burst assembly by setting length thresholds dynamically according to preset determination thresholds and estimated burst demand.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a fixed time threshold timer mechanism is used for convergence, then the convergence process is simple to implement, but the adaptability to traffic variations is weak and convergence precision is limited
Solution Approach 1:
The patent implements dynamic convergence threshold adjustment by replacing the fixed time threshold with a dynamically calculated threshold based on real-time traffic prediction. The convergence threshold is continuously updated according to predicted traffic patterns, allowing the system to adapt to varying traffic conditions while maintaining a relatively simple overall structure.
Solution Approach 2:
The patent introduces a feedback mechanism where the actual convergence results and traffic conditions are continuously monitored and fed back to adjust the convergence threshold. This closed-loop control enables the system to learn from past performance and optimize convergence parameters dynamically, improving adaptability without significantly increasing complexity.
2Adaptability or versatility
If a single burst as basic unit for threshold adjustment is used, then the adjustment process is simple, but the adaptive granularity is fixed and lacks flexibility
Solution Approach 1:
The patent segments the traffic flow into multiple granularities by introducing intermediate convergence units between single bursts and full convergence cycles. This multi-level segmentation allows threshold adjustments at different granularities, providing both fine-tuned adaptability and maintained convergence efficiency through hierarchical control.
Solution Approach 2:
The patent implements dynamic granularity selection where the basic adjustment unit can vary between single bursts, multiple bursts, or custom-sized groups based on traffic conditions. This dynamic adjustment of granularity allows the system to optimize between adaptability and efficiency depending on the specific traffic scenario.
3Measurement precision
If qualitative estimation by comparing burst length with threshold window is used, then the prediction mechanism is simple, but accurate traffic information cannot be provided and optimal dynamic adjustment is difficult to achieve
Solution Approach 1:
The patent replaces the simple mechanical comparison mechanism with a traffic prediction model that uses historical traffic data and statistical analysis. This substitution introduces more accurate traffic information estimation while keeping the overall system complexity manageable through efficient algorithms and data structures.
Solution Approach 2:
The patent performs preliminary traffic analysis and prediction before the actual convergence decision is made. By pre-calculating traffic patterns and predicting future convergence needs, the system achieves more accurate threshold adjustment without adding significant complexity to the real-time convergence process.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A service convergence method and system are provided. The method includes: receiving a total amount of services arriving at a convergence queue within a ended adaptive period, and calculating a predicted value of a total amount of the services arriving at the convergence queue within a next period; reading a length of the convergence queue, and obtaining an estimated burst demand of the next adaptive period; predicting a length threshold of the convergence queue used in the next adaptive period according to a preset burst demand determination threshold, and triggering one convergence when the length of the convergence queue reaches the predicted value of the length threshold or no convergence has been triggered within one period. Through the present invention, the adaptive precision of the convergence is increased, the flexibility of the convergence is enhanced, and the most appropriate adaptive granularity can be selected according to actual requirements, thereby optimizing the price performance ratio of the products.