Granular Policing via Multiple Arrival Curves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing policing algorithms in connection-oriented packet switched networks lack the ability to make granular policing decisions, as they rely on a single arrival curve, which fails to differentiate between cells that violate service level agreements by different factors, leading to unfair prioritization and potential discard of cells that should be treated differently based on their actual arrival characteristics.
Innovation Solution
A novel policing method that evaluates arriving cells' conformance by comparing their actual arrival characteristics against a family of ordered N arrival curves, allowing for the determination of which curves are violated and non-violated, thereby prioritizing non-conforming cells based on their degree of non-conformance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single arrival curve is used for policing decisions, then the policing process is simple, but granular policing decisions cannot be made and cells with different degrees of non-conformance cannot be differentiated
Solution Approach 1:
The patent segments the single arrival curve into multiple ordered arrival curves (first arrival curve, second arrival curve, etc.), each representing different tolerance thresholds. This segmentation allows the system to evaluate cells at multiple granularity levels, distinguishing between cells that violate only the second curve versus those that violate both first and second curves, thereby enabling precise granular policing decisions without excessive complexity.
2Device complexity
If a single tolerance threshold is used, then the policing algorithm is simple, but unfair prioritization occurs among non-conforming cells
Solution Approach 1:
The patent applies local quality by assigning different tolerance thresholds to different arrival curves. Each curve has its own specific tolerance level, allowing the system to apply locally optimized evaluation criteria to different cell arrival patterns. This ensures that cells are evaluated according to their specific deviation from expected behavior, achieving fair prioritization where cells violating stricter thresholds receive appropriate priority treatment.
3Measurement precision
If multiple arrival curves are used for granular policing, then accurate differentiation of non-conformance degree is achieved, but device complexity increases
Solution Approach 1:
The patent introduces an additional dimension of evaluation by ordering multiple arrival curves hierarchically (first, second, etc.). This dimensional ordering allows the system to evaluate cells along a spectrum of tolerance levels rather than a single threshold. The ordered structure provides a systematic framework for comparison, enabling precise differentiation of non-conformance degrees while managing complexity through structured organization of the evaluation criteria.
Data Source
AI summary
A method for granular policing includes a novel policing architecture which evaluates an arriving cell's conformance by comparing the actual arrival characteristic against a family or a set of expected arrival curves. In embodiments of the present invention, expected arrival characteristics based on a connection service level agreement (SLA) are represented by a family of N arrival curves. The degree of non-conformance of a cell is evaluated by inspecting, which m arrival curves were violated and which N-m arrival curves were non-violated by the arriving cells. Subsequently in context of discard, non-conforming tagged cells are further prioritized based on their degree of non-conformance. Thus cells associated with a higher degree of non-conformance shall be discarded before cells with a relatively lower degree of non-conformance.