Relocating Packet Processing Functions in Distributed Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current 3GPP EPC architecture leads to sub-optimal routing of data packets due to its anchored design, where all traffic for a user device must pass through a single Packet Data Network Gateway (PGW), resulting in increased latency even for devices close to each other, and lacks flexible means to insert new functions in the packet processing chain.
Innovation Solution
A method and arrangement for relocating packet processing functions in a distributed environment, allowing for optimal routing by determining and relocating functions based on contextual information such as device mobility, latency requirements, and network load, enabling flexible placement of functions across different network sites.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If all traffic for a user device passes through a single Packet Data Network Gateway (PGW) in the anchored EPC architecture, then network control and management are simplified, but routing latency increases and network efficiency deteriorates
Solution Approach 1:
The patent segments the packet processing functions from the anchored PGW architecture. Instead of all traffic passing through a single PGW, the processing chain is divided into multiple independent functions (F1, F2, F3) that can be distributed across different network nodes. This segmentation allows traffic to be routed directly through the distributed functions without converging at a single anchor point, thereby reducing latency while maintaining manageable complexity through modular design.
Solution Approach 2:
The patent transitions from a single-dimension anchored architecture (all traffic through PGW) to a multi-dimensional distributed architecture. Functions are placed at different network locations (local site, central office, regional data center, national data center) creating a hierarchical spatial distribution. This dimensional change enables traffic to flow through the most appropriate network node based on location and load, reducing latency without sacrificing control capabilities.
2Productivity
If packet processing functions are distributed across multiple network nodes, then routing efficiency and latency are improved, but network complexity and function management difficulty increase
Solution Approach 1:
The patent implements a universal function chain architecture where the same sequence of processing functions (F1, F2, F3) can be instantiated at multiple network locations. Each function instance is self-contained and can handle traffic independently. This universality allows the network to achieve high routing efficiency through distributed processing while managing complexity through standardized, repeatable function templates that can be deployed consistently across different nodes.
Solution Approach 2:
The patent incorporates feedback mechanisms where the controlling entity monitors traffic patterns, load conditions, and performance metrics across the distributed function instances. Based on this feedback, the system can dynamically adjust traffic routing decisions, relocate functions between nodes, or scale instances to optimize performance. This feedback loop enables the network to adapt to changing conditions, maintaining high routing efficiency while managing complexity through automated control.
3Stability of the object's composition
If the EPC architecture uses a fixed anchored design, then network stability and predictability are maintained, but flexibility to adapt to changing traffic patterns and insert new functions is reduced
Solution Approach 1:
The patent transforms the static anchored architecture into a dynamic distributed system. Function instances can be created, moved, scaled, and deleted based on real-time traffic patterns and network conditions. The controlling entity can dynamically relocate functions between network nodes (e.g., moving F2 from a local site to a central office during peak load) and dynamically insert new function instances into the processing chain. This dynamic capability maintains stability through controlled evolution while providing versatility for adapting to changing requirements.
4Loss of time
If packet processing functions are relocated frequently to optimize routing, then latency is reduced, but network resource overhead and operational expenses increase
Solution Approach 1:
The patent implements preliminary action by pre-positioning function instances at multiple network locations before traffic demands arise. The controlling entity proactively deploys function instances (F1, F2, F3) at strategic network nodes based on predicted traffic patterns and service requirements. This preliminary placement reduces the need for frequent reactive relocations, as functions are already positioned near potential traffic sources. The system performs preliminary load balancing and function distribution to minimize future relocation needs, thereby reducing latency while controlling resource overhead.
Data Source
AI summary
The invention relates to a method in an arrangement of a communication network, controlling a chain of functions for processing data packets of a flow associated with a device. The method comprises obtaining (610) an indication indicating that information associated with a context of at least one of the functions has been altered. The context is related to the flow and at least one of the communication network, the device, a route of the flow, and a source instance of each of the at least one function. The method further comprises determining (620) to relocate the at least one function for processing the data packets from the respective source instance to a respective target instance based on the obtained indication. The method also comprises relocating (630) the at least one function for processing the data packets to the respective target instance.