Multi-Level Network Capacity Allocation for Service Flow Prioritization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network congestion management methods often result in a poor user experience due to predefined limits and adaptive queue management, which may discard packets and lead to inefficient data transmission.
Innovation Solution
A multi-level allocation of network capacity across various service flow types, where minimum commitments are ensured for each service flow, and variable allocations are made based on priority, with policies that slow down certain service flows during congestion and adapt based on usage patterns, using a scheduler to manage thousands of queues and allocate capacity across multiple cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If adaptive queue management is used to discard packets during congestion, then network load is reduced to match capacity, but user experience deteriorates due to packet loss
Solution Approach 1:
The patent segments network traffic into multiple service flows with different priority levels (e.g., high-priority, medium-priority, low-priority). During congestion, the system selectively manages different service flow types rather than uniformly discarding packets. This segmentation allows the network to preserve user experience for critical services while managing overall capacity utilization.
Solution Approach 2:
The patent applies different quality of service (QoS) policies to different service flows based on their characteristics and priority. High-priority service flows receive preferential treatment with guaranteed bandwidth and lower packet loss, while lower-priority flows are more aggressively managed during congestion. This local quality approach ensures that user experience is maintained for important services while still achieving capacity management.
2Device complexity
If predefined limits are imposed on subscriber traffic, then network capacity allocation is simplified, but allocation efficiency deteriorates during congestion
Solution Approach 1:
The patent implements dynamic capacity allocation that adjusts in real-time based on network conditions, service flow priority, and subscriber behavior. Rather than using static predefined limits, the system continuously monitors congestion levels and dynamically re分配s capacity among service flows. This dynamic approach maintains manageable complexity while significantly improving transmission efficiency during congestion events.
Solution Approach 2:
The patent incorporates feedback mechanisms that monitor network congestion, service flow performance, and subscriber usage patterns. This feedback is used to continuously adjust capacity allocation decisions, ensuring that the system responds to changing conditions. The feedback loop enables the network to maintain optimal efficiency without requiring complex manual configuration.
3Productivity
If uniform packet discarding is applied during congestion, then network load is controlled, but data transmission efficiency deteriorates due to loss of valuable data
Solution Approach 1:
The patent applies differential packet management based on service flow priority and data importance. Instead of uniform discarding, the system selectively manages packets from different service flows, preserving high-priority data while managing lower-priority traffic. This approach controls network load while minimizing loss of valuable data.
Solution Approach 2:
The patent converts the harmful effect of congestion into a beneficial opportunity for optimization. By analyzing which service flows and data types are most affected by congestion, the system learns to prioritize them in future allocation decisions. This transforms the problem of data loss during congestion into a benefit of improved long-term allocation efficiency and reduced future losses.
Data Source
AI summary
Disclosed herein are systems and methods for allocating network capacity over a communication channel of a network. The systems and methods determine a transmission profile for each of a plurality of service flow types. The systems and methods then iteratively perform the following steps for allocating network capacity: selecting, for each service flow type, the network capacity allocation parameters in each service flow type's transmission profile associated with a current network capacity allocation cycle; determining amounts of data to transmit for each of the plurality of service flow types based at least in part on the selected network capacity allocation parameters; and transmitting, over the communication channel, the determined amounts for each of the plurality of service flow types for the current network capacity allocation cycle.