PDN-GW Pool Resilience via N+M Standby Activation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3GPP LTE specifications do not adequately address geographic redundancy and in-service maintenance for Evolved Packet Core (EPC) components like PDN-GWs and SGWs, leading to inefficient use of resources and potential service disruptions.
Innovation Solution
Implementing an N+M pooled resiliency scheme across PDN-GW and SGW pools, where each pool element stores UE session information and allows for geographic redundancy and in-service maintenance without impacting ongoing sessions, using a method that smoothly transitions between active and standby nodes to ensure continuous service.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 1+1 network level redundancy solutions are implemented, then service reliability is improved, but resource utilization deteriorates due to 50% capacity being used only for redundancy
Solution Approach 1:
The patent merges multiple active gateways into a pooled architecture where N gateways share M backup resources. Instead of dedicating separate backup resources to each gateway (1+1 redundancy), the system combines backup functions across the entire pool, allowing resources to be dynamically shared and utilized by multiple gateways simultaneously.
Solution Approach 2:
The backup gateway in the N+M pooled architecture serves multiple functions: it can back up any of the N active gateways and can be dynamically assigned to different active gateways based on failure patterns and load conditions. This universal backup capability eliminates the need for dedicated backup resources for each active gateway.
2Reliability
If geographic redundancy is implemented, then service continuity is improved during failures, but system complexity increases due to distributed deployment requirements
Solution Approach 1:
The patent segments the gateway pool into N active gateways and M backup gateways that can be geographically distributed. Each gateway is an independent unit that can be deployed in different locations, allowing the system to achieve geographic redundancy through modular segmentation rather than complex integrated systems.
Solution Approach 2:
The MME (Mobility Management Entity) acts as an intermediary that manages the pooling architecture, handling the complexity of resource allocation, failover coordination, and session information management. This centralizes the control logic and simplifies the overall system architecture by having a dedicated management entity rather than distributed complex coordination between all gateways.
3Reliability
If dedicated backup gateways are assigned to each active gateway, then failover reliability is improved, but cost increases due to unnecessary redundancy
Solution Approach 1:
Instead of providing full backup capacity for each active gateway (which would require N backup gateways for N active gateways), the system uses partial backup action where M backup gateways serve N active gateways with M < N. This partial redundancy is sufficient to handle failures while reducing the total number of gateway resources required.
Solution Approach 2:
The system allows backup gateways to be dynamically recovered and reassigned. When a backup gateway is not currently backing up any active gateway, it can be recovered and assigned to a different active gateway that experiences failure, maximizing the utilization of backup resources and reducing the total number of gateways needed.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
Embodiments of the invention include a method for providing UE session resilience performed in a first PDN-GW that is coupled to a second PDN-GW, which are both in a PDN-GW pool. The method provides UE session resilience by allowing the first PDN-GW to provide connectivity for UE sessions previously serviced by the second PDN-GW after the second PDN-GW becomes non-operational. The first PDN-GW recognizes that the second PDN-GW failed and then activates a plurality of standby UE sessions. Each standby UE session is a backup UE session corresponding to a previously active UE session serviced on the second PDN-GW. Each standby UE session is associated with a UE device and a network resource identifier of an APN slice. The first PDN-GW transmits a message to a SGW that is servicing the UE sessions that indicates that the SGW should direct traffic previously bound for the second PDN-GW to the first PDN-GW.