Serverless Core Network Architecture for 5G Resource Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G networks consume unnecessary resources due to continuously running services, and routing signaling traffic between home and visited networks can be challenging, leading to delays and inefficient resource usage.
Innovation Solution
Implementing a serverless Core network architecture that pre-distributes software images of services to roaming partners and instantiates them only when needed, allowing services to execute closer to the user's location, with mechanisms for separate data and software management and automatic service triggering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If services are continuously running in current 5G networks, then service availability is maintained, but resource consumption increases unnecessarily
Solution Approach 1:
The patent implements serverless computing where services are executed periodically or on-demand rather than continuously. Services are instantiated only when triggered by specific events or requests, and terminated when not needed, converting continuous operation into periodic or event-driven execution. This reduces resource consumption while maintaining service availability when required.
Solution Approach 2:
The system employs automatic service triggering and management mechanisms where services self-start when needed and self-terminate when not needed. The network automatically manages service lifecycle based on detected triggers or requests, eliminating the need for continuous manual provisioning while ensuring services are available when required.
2Device complexity
If services are routed between home and visited networks, then centralized control is maintained, but signaling delays and resource inefficiency occur
Solution Approach 1:
The patent distributes service execution to local visited networks where users are currently located, rather than routing all signaling through home networks. Services are instantiated and executed locally in the visited network, reducing signaling distance and delay while maintaining centralized service management through the service management function. This localizes service delivery to improve responsiveness.
Solution Approach 2:
The system segments service management into two parts: centralized service definition and management (in home network) and distributed service execution (in visited networks). This segmentation allows centralized control of service policies while enabling local execution to reduce signaling delays. Services are divided between management plane and execution plane across different network locations.
3Productivity
If services are pre-distributed to roaming partners, then service deployment speed increases, but initial resource allocation complexity increases
Solution Approach 1:
The patent pre-distributes service software images or containers to roaming partner networks in advance, so that when a user roams or a service is needed, the service can be instantly instantiated without downloading or provisioning delays. This preliminary distribution of service packages accelerates deployment while the actual service instantiation remains lightweight and trigger-based.
Solution Approach 2:
The system creates and distributes copies of service software images or containers to multiple roaming partner networks. These copies can be independently instantiated without affecting the original service definition. This copying mechanism enables fast local deployment while the central system maintains the master service definition, managing complexity through replication rather than complex allocation logic.
Data Source
Figure 1A~1B
Figure 2
Figure 3
AI summary
Methods for provisioning a service in a first network and providing the service by an entity in a second network are disclosed. The methods comprising maintaining in a software repository in the first network, one or more software images (I1, I2,...Ii) of corresponding one or more services and where each service is identified by a service identifier (Si) and storing configuration information related to the one or more services in a database repository of the first network. The method of providing the service in the second network further comprises receiving and storing one or more software images (I1, I2,...) for one or more services of the first network and where each service is identified by a service identifier (Si) and upon detecting a trigger indicating a request for a service of the one or more services, instantiating the software image of the service in the serving network.