5G Edge Service Routing With Weighted Anycast Server Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G edge computing environments face challenges in selecting the best application server for user equipment (UE) due to reliance on proximity-based metrics, which can lead to overloaded servers and increased latency, especially when UE moves between 5G sites.
Innovation Solution
A weighted approach is used to determine the best path for packet transmission by considering network latency, load, capacity, and preference of application servers, using Border Gateway Protocol (BGP) messages to gather metrics and calculate a cost function for selecting optimal egress routers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If proximity-based metrics are used to select application servers, then network latency is reduced, but server overload occurs and reliability deteriorates
Solution Approach 1:
The patent changes the selection parameters from simple proximity-based metrics to a composite cost function that includes multiple weighted parameters: network latency (D), load measurement (L), capacity index (β), and network preference (γ). This allows the system to balance multiple factors rather than optimizing for latency alone, preventing server overload while maintaining low latency performance.
Solution Approach 2:
The patent implements a feedback mechanism where egress routers continuously report load measurements, capacity indexes, and network preferences to ingress routers. This real-time feedback enables dynamic adjustment of the cost function weights and parameters, allowing the system to adapt to changing network conditions and server states, thereby maintaining reliability while optimizing latency.
2Reliability
If multiple application servers with the same anycast address are used, then service redundancy is improved, but traffic distribution becomes unbalanced
Solution Approach 1:
The patent applies local quality by allowing different application servers to have different local characteristics (load measurements, capacity indexes, network preferences) while maintaining the same anycast address. The cost function incorporates these local differences through weighted parameters, enabling differentiated treatment of each server based on its current state and capabilities, thus achieving balanced traffic distribution across redundant servers.
3Device complexity
If simple routing based on anycast address is used, then device complexity is reduced, but network delivery optimization is insufficient
Solution Approach 1:
The patent introduces an intermediary cost function that mediates between simple anycast routing and complex multi-criteria optimization. The cost function C(D, L, β, γ) acts as a mathematical intermediary that consolidates multiple complex parameters into a single comparable value, allowing routers to make optimized decisions without dealing with the full complexity of multiple simultaneous parameters. This maintains relative simplicity while achieving comprehensive optimization.
Data Source
AI summary
A method and apparatus for optimizing network delivery of fifth generation (5G) edge application services by weighting network latency along with a load, capacity, and preference of an application server to determine where to transmit a new flow of packets. By using the weighted approach, network traffic can be equalized among multiple instances (e.g., applications) with the same address.