Network Function Virtualization for Dynamic Capacity Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network infrastructures face challenges in scalability and flexibility to handle varying data volumes and loads, leading to inefficiencies in capacity utilization and management, especially in sparsely populated areas.
Innovation Solution
A service-oriented architecture that decomposes network entities into atomic services, allowing for flexible deployment and scaling of services based on demand, with the ability to add or remove services and servers dynamically to match network load, optimizing resource usage and capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed capacity network infrastructures are implemented to meet peak load demands, then network capacity is sufficient during peak periods, but capacity is wasted when network load is light
Solution Approach 1:
The patent implements dynamic capacity adjustment by allowing network functions to be migrated between physical and virtualized network functions based on traffic conditions. The system dynamically scales network capacity by activating or deactivating virtual network function instances, enabling the network to adapt its capacity to match actual demand rather than maintaining fixed peak-capacity infrastructure.
2Adaptability or versatility
If network infrastructures are expanded globally to provide internet service to unconnected residences, then network coverage and accessibility are improved, but infrastructure cost and complexity increase
Solution Approach 1:
The patent creates a universal network function virtualization platform that can serve multiple network functions and services through a single infrastructure. Virtual network functions can be deployed across different geographic locations and hardware platforms, providing multi-functional capability that reduces overall infrastructure complexity while expanding coverage.
Solution Approach 2:
The patent uses virtualization to create software copies of network functions that can be replicated and distributed across multiple physical locations. Instead of deploying unique hardware for each network function at each location, the system copies virtual network function instances, reducing infrastructure complexity while enabling global expansion.
3Stability of the object's composition
If traditional network architecture with fixed network entities is used, then network stability is maintained, but flexibility to handle varying data volumes and loads is reduced
Solution Approach 1:
The patent segments traditional fixed network entities into independent, modular network functions that can be individually managed, deployed, and scaled. By separating network functions from hardware platforms and enabling independent virtualization of each function, the system maintains stability through standardized interfaces while gaining flexibility to adjust individual function capacities based on demand.
Solution Approach 2:
The patent transforms static network architecture into a dynamic system where virtual network functions can be migrated, scaled, and reconfigured based on real-time traffic conditions. The system maintains stability through consistent service interfaces while enabling flexible adaptation to varying data volumes through dynamic resource allocation and load balancing across virtual function instances.
Data Source
AI summary
Particular embodiments may detect, by a core network, a change in network traffic types from a first network traffic type to a second network traffic type. The core network includes one or more network functionality components. Each of the one or more network functionality components is decomposed into multiple service types. The core network may determine several service instances for deployment in response to the change in the network traffic types. Each of the service instances may belong to one of the multiple decomposed service types. The core network may deploy several service instances to one or more server machines of the core network according to a decomposed service type of a respective service instance.