Leaky Bucket Traffic Limitation via Sub-bucket Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data network traffic limitation methods using leaky buckets fail to distinguish and flexibly manage packets with different priorities, leading to inadequate Quality of Service (QoS) and user experience.
Innovation Solution
Divide the overall leaky bucket into sub-buckets based on packet priority parameters and ratios, allowing for dynamic token allocation according to weighted priorities, ensuring that packets with higher priorities receive more tokens and overrunning tokens are shared among sub-buckets, thereby enabling flexible traffic limitation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single leaky bucket is used for traffic limitation, then the device complexity is reduced, but the ability to distinguish and manage packets with different priorities deteriorates
Solution Approach 1:
The single leaky bucket is segmented into multiple sub-buckets, where each sub-bucket corresponds to a specific priority level. This segmentation allows the system to track and manage tokens for different priority packets separately, enabling priority-based traffic differentiation while maintaining the overall simplicity of the leaky bucket approach.
Solution Approach 2:
Each sub-bucket is assigned local quality characteristics specific to its priority level, such as different token generation rates or different thresholds for packet acceptance. This allows high-priority packets to receive preferential treatment (e.g., higher token accumulation rates) while low-priority packets are more strictly limited, achieving flexible QoS management.
2Ease of operation
If tokens are uniformly allocated to all packets, then the ease of operation is maintained, but the Quality of Service for different priority packets deteriorates
Solution Approach 1:
The token allocation process applies different rules to different priority levels. High-priority packets may receive tokens at a higher rate or with lower thresholds, while low-priority packets receive tokens more conservatively. This local differentiation improves QoS for critical traffic while maintaining automated operation without complex manual intervention.
Solution Approach 2:
The token allocation rates and thresholds for each sub-bucket can be dynamically adjusted based on network conditions and priority requirements. This allows the system to adapt to changing traffic patterns and QoS demands while maintaining ease of operation through automated control mechanisms.
3Device complexity
If the leaky bucket volume is fixed, then the device complexity is reduced, but the flexibility in bandwidth allocation for different priorities deteriorates
Solution Approach 1:
The fixed overall bucket volume is segmented into multiple sub-buckets with individually configurable volumes. Each sub-bucket can be allocated a specific portion of the total capacity based on priority requirements, allowing flexible bandwidth distribution while maintaining the simplicity of fixed-volume management at the aggregate level.
Solution Approach 2:
The volume parameters of individual sub-buckets can be changed to reflect different bandwidth allocation requirements for various priority levels. This allows the system to adapt bandwidth allocation dynamically by adjusting sub-bucket volumes without changing the overall fixed-volume architecture, maintaining device simplicity while achieving allocation flexibility.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The disclosure discloses a method for traffic limitation by a leaky bucket, including that: during traffic-limitation of a group of packets sharing a same leaky bucket and having separate weighted priorities, the overall leaky bucket is divided into multiple sub-buckets according to priority parameters of the packets including a number of the weighted priorities and ratios between the weighted priorities; after packet information is received, parameters relevant to traffic-limitation by a leaky bucket is acquired; a total number of tokens to be added to the overall leaky bucket is calculated; token-addition is performed on the sub-buckets according to a state of the overall leaky bucket after addition of the total number of tokens and the weighted priorities corresponding to the sub-buckets; a packet is forwarded, or discarded, or marked according to a length of the packet and a number of tokens in a sub-bucket corresponding to a weighted priority, wherein in performing token-addition on the sub-buckets, after token-addition is performed on all the sub-buckets, any overrunning token is allocated to another sub-bucket not overrun with tokens. The disclosure further discloses a device for traffic limitation by a leaky bucket. With the disclosure, it is possible to limit packet traffic flexibly according to weighted priorities and improve resource utilization.