TCP Traffic Control via RED and Window Size Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network congestion management techniques, such as tail drop and Active Queue Management (AQM), face challenges like global synchronization, network underutilization, and oscillation in queue size, especially in satellite networks with varying bandwidth and latency.
Innovation Solution
A method combining Random Early Detection (RED) with window size adjustments, where a simple moving average of packet dropping probabilities is calculated to adjust the advertised window size based on tolerance factors, allowing for more responsive and stable traffic control without unnecessary packet dropping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If random early detection is used to probabilistically drop packets based on queue size, then fast response and autonomous rate control are achieved, but oscillation in queue size and unwanted larger packet dropping rates occur
Solution Approach 1:
The patent implements feedback control by continuously monitoring the queue size and adjusting the packet dropping probability accordingly. The system calculates the average queue size over a time window and uses this feedback to modulate the dropping rate, creating a closed-loop control system that responds to network conditions while dampening oscillations through proportional control.
Solution Approach 2:
The patent introduces dynamic adjustment of the dropping probability parameter based on current network conditions. Instead of using a fixed dropping rate, the system dynamically modifies the probability parameter according to the measured queue size, allowing the control mechanism to adapt to changing traffic patterns and maintain stability across varying load conditions.
2Productivity
If the queue size threshold for triggering TCP congestion control is set high, then network underutilization is reduced, but unnecessary packet delay occurs when the network is in equilibrium
Solution Approach 1:
The patent applies preliminary action by proactively dropping packets before the queue reaches its maximum threshold. Instead of waiting for the queue to fill up (reactive tail drop), the system preemptively removes excess packets when the queue size exceeds a lower threshold, preventing both congestion and the associated delays while maintaining better network utilization.
Solution Approach 2:
The patent changes the critical parameter from a fixed queue size threshold to a dynamic dropping probability that varies with queue size. This parameter transformation allows for smoother control of traffic flow, enabling the system to maintain optimal utilization by adjusting the effective threshold behavior continuously rather than switching at a single fixed point.
3Reliability
If the queue size threshold is set low to prevent congestion, then network recovery is faster, but the processing rate drains the queue before TCP flows recover, causing network link underutilization
Solution Approach 1:
The patent implements periodic action through cyclic modulation of the dropping probability. The system applies dropping in periodic cycles based on queue size thresholds, allowing the queue to be drained during high-load periods and naturally refill during lower-load periods, creating a rhythmic control pattern that prevents both persistent congestion and complete queue drainage.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for controlling data traffic with random early detection and window size adjustments including performing random early detection on incoming data packets, calculating a simple moving average of packet dropping probabilities for the data packets as calculated when performing random early detection, decreasing an advertised window size if the simple moving average is greater than a probability target plus a tolerance factor, increasing the advertised window size if the simple moving average is less than the probability target minus a tolerance factor, and not adjusting the window size if the simple moving average is not greater than a probability target plus a tolerance factor and not less than a probability target minus a tolerance factor.