Network Congestion Control via Predictive Parameter Configuration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing adaptive communication techniques, such as IEEE 802.3x PAUSE and Priority Flow Control, are inadequate in managing rapid changes in user data throughput and maintaining quality of service (QoS) across varying network conditions, particularly due to limitations in buffer management and link utilization.
Innovation Solution
A method and system that collect network topology information and parameters to predict future link conditions, compute optimal congestion mitigation parameters, and configure each node's specific congestion mechanisms to dynamically adjust transmission parameters, ensuring lossless behavior and maximum network utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If simple pause or flow control mechanisms are used, then device complexity is reduced, but productivity deteriorates due to insufficient handling of rapid changes in user data throughput
Solution Approach 1:
The system performs preliminary actions by collecting network topology information and computing predicted network conditions before congestion occurs. This allows the system to proactively configure optimal congestion mitigation parameters and dynamically adjust transmission parameters in advance, enabling effective handling of rapid throughput changes without requiring complex reactive mechanisms.
Solution Approach 2:
The system implements feedback by continuously collecting network topology information from nodes, computing predicted network conditions based on this data, and using the predictions to dynamically adjust transmission parameters. This closed-loop feedback mechanism enables the system to adapt to changing network conditions and maintain productivity without increasing device complexity.
2Reliability
If buffer threshold is set low to prevent buffer overflow, then reliability is improved, but productivity deteriorates due to reduced link utilization
Solution Approach 1:
The system computes predicted network conditions and optimal congestion mitigation parameters before buffer overflow occurs. By proactively configuring transmission parameters based on predictions, the system can maintain higher buffer thresholds without risking overflow, thus preserving link utilization while ensuring lossless behavior.
Solution Approach 2:
The system dynamically adjusts transmission parameters based on predicted network conditions rather than using fixed buffer thresholds. This dynamic adaptation allows the system to optimize the balance between reliability and productivity by adjusting parameters in real-time according to actual network state, preventing both buffer overflow and excessive conservatism.
3Reliability
If adaptive coding and modulation schemes are used, then reliability is improved by maintaining service availability under varying conditions, but productivity deteriorates due to detection delay between network condition changes and communication adjustments
Solution Approach 1:
The system collects network topology information and computes predicted network conditions in advance, before actual congestion or performance degradation occurs. This preliminary computation enables the system to proactively adjust communication parameters, eliminating detection delay and maintaining both reliability and rapid response to network changes.
Data Source
AI summary
The present invention relates to a method for controlling traffic congestion in a network, the network comprising a plurality of nodes including a source node, intermediary nodes and a destination node. In one embodiment, this can be accomplished by collecting network topology information and various parameters from each nodes in the network, storing the collected information and parameters in a database, computing network statistics from at least one of the contemplated collected information thereby reflecting the probabilities of changing modulation at a specific state, computing optimal parameters of specific congestion mitigation mechanisms associated with the plurality of nodes and configuring each node's specific congestion mitigation mechanism, wherein for each mechanism the method calculates the optimal parameter configuration per node.