Feedback-Driven Queue Admission Control for Data Loss and Delay
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional queuing systems face issues with data blocking, causing delays and degradation when queues are full, or data loss when they are non-blocking, leading to inefficiencies and errors.
Innovation Solution
Implementing intelligent queuing with feedback-driven admission controls that dynamically change between blocking and non-blocking states based on transmission success rates, using exponential decay filters to adjust queue states based on success and failure conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a queue is set to blocking state, then data loss is reduced, but waiting time increases
Solution Approach 1:
The queue dynamically switches between blocking and non-blocking states based on transmission success rates. When success rate exceeds threshold, queue becomes non-blocking to reduce waiting time; when success rate drops below threshold, queue becomes blocking to prevent data loss. This dynamic adaptation resolves the contradiction by adjusting queue behavior according to real-time transmission conditions.
Solution Approach 2:
The system continuously monitors transmission success rates and uses this feedback to adjust queue state. The exponential decay filter processes success/failure signals to determine when to switch queue states. This feedback mechanism ensures the queue adapts to changing transmission conditions, balancing data loss prevention with waiting time reduction.
2Loss of time
If a queue is set to non-blocking state, then waiting time is reduced, but data loss increases
Solution Approach 1:
The queue transitions to non-blocking state only when transmission success rate exceeds the threshold, indicating reliable downstream system performance. This dynamic state change allows the system to reduce waiting time during favorable conditions while maintaining data integrity through state-aware admission control.
Solution Approach 2:
The system changes the queue's operational parameter (blocking vs. non-blocking) based on the transmission success rate parameter. When success rate crosses the threshold boundary, the queue parameter switches accordingly, enabling the system to optimize between waiting time and data loss based on current transmission conditions.
3Reliability
If transmission retries are increased, then data loss is reduced, but transmission time increases
Solution Approach 1:
The system performs preliminary admission control by checking queue state before allowing data entry. When queue is in non-blocking state with high success rate, data is transmitted immediately without unnecessary retries. When success rate drops, the system prevents new data entry proactively, avoiding wasted retry time and reducing overall transmission time while maintaining reliability.
Solution Approach 2:
The exponential decay filter continuously processes transmission success/failure feedback to adjust the effective transmission strategy. By weighting recent successes more heavily and decaying older information, the system adapts retry behavior to current conditions, reducing unnecessary retries when success rate is high and preventing data loss when success rate drops.
Data Source
AI summary
Novel tools and techniques are provided for implementing intelligent queuing with feedback-driven admission controls. In various embodiments, a computing system may send, to a downstream system, first data among a plurality of data that is stored in a queue; may determine whether or not the first data has been successfully transmitted to and received by the downstream system; based on a determination that the first data has not been successfully received by the downstream system and when the first queue is in a first state, may analyze a number of successful transmission attempts and a number of unsuccessful transmission attempts within a first predetermined window, to determine whether a first trigger event has occurred; and if so, may change a state of the queue from the first state to a second state, the first and second states of the queue defining how data is enqueued by the queue.