Network Load API Predicting Congestion for Packet Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network congestion APIs are primarily backward-looking, failing to provide real-time predictions of long duration congestion events, which disrupts the continuous transmission of multimedia content over mobile wireless cellular networks, leading to poor user experience for applications that rely on real-time delivery of communication packets.
Innovation Solution
A network load API that includes indicators for long duration congestion events, using statistical methods to determine stability factors and predict future congestion levels, allowing applications to schedule packet transmissions and adapt content delivery accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional backward-looking network congestion APIs are used, then current network status can be monitored, but real-time prediction of long duration congestion events is not available, causing packet transmission delays
Solution Approach 1:
The patent applies preliminary action by using statistical methods to predict future congestion levels before they actually occur. The system calculates stability factors from historical network load data and generates predictions about upcoming congestion events, allowing applications to proactively adjust their packet transmission schedules and avoid delays caused by congestion before it happens.
2Productivity
If network congestion prediction is implemented, then applications can schedule transmissions to avoid congestion, but the API complexity increases with statistical methods and predictive analytics
Solution Approach 1:
The patent uses an intermediary approach by introducing a network load API that acts as a mediator between the complex statistical prediction system and the applications. The API handles the complexity of statistical calculations, stability factor determination, and congestion prediction internally, while presenting a simplified interface to applications that only needs to query for predicted congestion levels and adjust transmissions accordingly.
3Measurement precision
If real-time network utilization data is collected and analyzed, then accurate congestion predictions can be made, but the processing requirements and data collection overhead increase
Solution Approach 1:
The patent applies partial action by collecting and analyzing only the specific network load parameters necessary for congestion prediction, rather than processing all possible network data. The system focuses on gathering relevant utilization metrics from network elements and applies statistical methods selectively to generate stability factors and predictions, reducing unnecessary processing overhead while maintaining prediction accuracy.
Data Source
AI summary
Reception of network load data is disclosed. For instance, the network load data can provide indication of a utilization level extant in a wireless cellular network. The systems and methods, as a function of the utilization level, determine a congestion metric that indicates a level of congestion determined to have been experienced by a communication packet using the wireless cellular network device. Further, the disclosed systems and methods schedules transmission of communication packets to an end user device via the wireless cellular network device, as a function of the determined congestion metric.