Adaptive Packet Distribution Across Network Function Instances and LAGs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional load-balancing systems are ineffective in distributing packets across network function instances and intermediate link aggregation groups (LAGs), leading to inefficient traffic distribution and increased operational and capital costs due to overburdened network devices.
Innovation Solution
A method and system for adaptive multi-criteria packet distribution that determines paths based on link capacity and network function instance capacity, using a controller to dynamically adjust load-balancing decisions and build rule tables that optimize traffic distribution across both network function instances and LAGs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional load-balancing systems are used to distribute packets, then traffic distribution is simplified, but network devices become overburdened and efficiency decreases
Solution Approach 1:
The patent implements dynamic load-balancing that adapts to current network conditions by continuously monitoring link capacity and network function instance capacity, adjusting packet distribution paths in real-time based on variance thresholds and available resources
Solution Approach 2:
The system changes distribution parameters dynamically by evaluating link capacity, network function instance capacity, and variance metrics to determine optimal packet routing decisions, switching between different distribution strategies based on current state
2Power
If more network devices and links are deployed to handle traffic, then network capacity increases, but operational and capital costs increase
Solution Approach 1:
The system enables network resources to serve themselves optimally by automatically monitoring capacity metrics and making intelligent distribution decisions without external intervention, maximizing utilization of existing devices and links
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor link capacity and network function instance capacity, using this information to dynamically adjust packet distribution and optimize resource utilization
3Device complexity
If packets are distributed based on only one criterion (e.g., link capacity), then distribution decisions are simple, but overall network optimization is insufficient
Solution Approach 1:
The patent segments the load-balancing decision into two distinct modules: one handling link capacity-based distribution and another handling network function instance capacity-based distribution, with a controller that switches between them based on variance thresholds
Data Source
AI summary
The disclosure is directed at a system and method for packet distribution in a network. After receiving a data packet, at least one network function associated with the packet, at least one link aggregation group (LAG) associated with the packet, and a variance associated with a plurality of links within the at least one LAG are determined. A determination as to whether the variance is above a pre-determined threshold is then performed and if the variance is above the threshold, a path for the packet based on capacity associated with each of the plurality of links is determined. Otherwise a path based on capacity of each of a plurality of network function instances associated with the at least one network function is determined.