Dynamic Core Sharding for Low-Latency 5G Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The fixed locations of core data centers in 5G networks limit flexibility and efficiency in routing communications, leading to increased latency and communication errors due to indirect routes and the need for additional hardware deployment.
Innovation Solution
Implementing ephemeral core shards in data centers outside the fixed network core, allowing dynamic provisioning and deprovisioning of core services to create shorter, more direct routes using cloud-based infrastructure across multiple cloud providers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If fixed core data centers are used in 5G networks, then network stability is maintained, but routing flexibility is limited and latency increases
Solution Approach 1:
The patent implements dynamic core sharding that allows the network to transition from static fixed data centers to dynamic ephemeral shards. Core services are divided into multiple shards that can be dynamically created, moved, and destroyed across different data centers based on real-time network conditions, traffic patterns, and service requirements. This enables the network to adapt routing paths dynamically while maintaining service stability through coordinated state management.
Solution Approach 2:
The core network services are segmented into multiple independent shards instead of relying on a single fixed core data center. Each shard handles specific core services and can be independently provisioned, managed, and routed. This segmentation allows different parts of the core network to be optimized independently, improving overall routing flexibility and communication speed while maintaining system stability through distributed architecture.
2Quantity of substance
If more hardware is deployed to expand network coverage, then bandwidth capacity increases, but infrastructure complexity and cost increase
Solution Approach 1:
The ephemeral core shards can be deployed across existing data center infrastructure from multiple cloud providers, making the same hardware resources serve multiple network functions and multiple core services simultaneously. A single data center can host multiple core shards for different services or different geographic regions, reducing the need for dedicated hardware deployments and lowering infrastructure complexity while maintaining or expanding network capacity.
Solution Approach 2:
Instead of physically deploying more hardware to expand capacity, the system changes operational parameters by dynamically provisioning and deprovisioning core shards based on demand. The network capacity is expanded through virtualization and software-defined networking parameters rather than physical hardware additions, allowing flexible capacity adjustment without proportional increases in infrastructure complexity.
3Area of stationary object
If indirect routing paths are used through fixed data centers, then network coverage is extended, but latency and communication errors increase
Solution Approach 1:
The patent introduces ephemeral core shards as intermediary nodes between edge networks and fixed data centers. These intermediary shards are dynamically positioned closer to user traffic sources and can handle local core services, reducing the need for long-distance routing through distant fixed data centers. This intermediary layer maintains extended network coverage while significantly reducing latency by minimizing the physical distance and number of hops communications must traverse.
Data Source
AI summary
Systems, methods, and devices perform core services on a 5G data and telephone network. A first edge data center of a first cloud network receives a communication from user equipment (UE). The communication is directed to another UE. A potential core shard from a plurality of potential core shards is selected to route the communication. The potential core shard includes a first instance at a first data center of a cloud network and a second instance at a second data center of the cloud network. The first instance is provisioned at the first data center, and the second instance is provisioned at the second data center. The communication is routed from an edge service to the first instance over a first communication channel, from the first instance to the second instance over the core communication channel, and from the second instance to the second UE over a third communication channel.