Dynamic WAN Latency Weighting for Bufferbloat Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques fail to effectively improve network performance when a network device has multiple WANs, as they are unable to adequately address the issue of bufferbloat and resulting latency in such multi-WAN environments.
Innovation Solution
The method involves establishing multiple connections between network devices, determining non-congesting latency for each connection, assigning weightings based on latency, and dynamically adjusting these weightings when connection performance deteriorates, thereby optimizing data packet routing across WANs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple WAN connections are used to send data packets, then network bandwidth and throughput are improved, but network latency increases due to bufferbloat and congestion on individual connections
Solution Approach 1:
The system dynamically adjusts the weighting of each WAN connection based on real-time latency measurements. The weighting factors are continuously updated to reflect current network conditions, allowing the system to adapt to changing congestion levels and minimize latency while utilizing multiple WAN connections for throughput.
Solution Approach 2:
The system changes the parameter of weighting assignment from static to dynamic based on latency performance. By monitoring latency metrics and adjusting weighting parameters accordingly, the system optimizes the distribution of data packets across multiple WAN connections to balance throughput and latency.
2Reliability
If buffer sizes are increased to handle congestion, then data packet loss is reduced, but latency increases significantly
Solution Approach 1:
The system segments the network traffic distribution across multiple WAN connections rather than relying on a single large buffer. By dividing traffic flow and dynamically routing packets through multiple paths based on latency-weighted decisions, the system avoids the latency penalty of large buffers while maintaining reliable packet delivery.
3Device complexity
If static weighting is assigned to multiple WAN connections, then system complexity is reduced, but network performance cannot be optimized under varying congestion conditions
Solution Approach 1:
The system implements feedback by continuously measuring latency on each WAN connection and using this information to adjust weighting assignments. This closed-loop control enables the system to automatically optimize network performance across multiple WANs without manual intervention, balancing complexity and adaptability.
Data Source
AI summary
The present invention discloses methods and systems for reducing network latency. A first network device establishes a plurality of connections with a second network device. After that, determining non-congesting latency of each of the plurality of connections. Assigning a weighting to each of the plurality of connections. Decreasing the weighting of a connection when the performance of the connection deteriorated according to a first criteria. The first network device may perform weight decreasing again after a time interval. Last, sending data packets through the plurality of connections according to the weightings.