Distributed Data Mesh for Low-Latency 5G Data Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile networks face challenges in managing vast amounts of data generated across network infrastructure components, leading to data loss, increased resource expenditure, and issues with data access and storage due to the complexity of virtualization and cloud-roaming, exacerbated by 5G networks.
Innovation Solution
A distributed data mesh system is implemented to manage and organize data by identifying data domains, storing data in localized data pools, and using a data catalog to ensure data access without moving data across domains, thereby preventing data loss and reducing redundant storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data is centralized in a single data lake, then data access is simplified, but data latency increases and architecture complexity increases
Solution Approach 1:
The patent divides the centralized data lake into multiple distributed data pools organized by data domains (e.g., subscriber data, billing data, location data). Each data pool stores data locally near its source generators, enabling fast access without centralized bottlenecks while maintaining simplified access through the data mesh framework.
Solution Approach 2:
The patent implements local data storage by placing data pools close to their respective data source generators (network elements, applications). This local quality approach reduces data latency by eliminating long-distance data transfers while maintaining ease of access through standardized data product interfaces.
2Reliability
If data is stored in multiple locations, then data redundancy increases, but data loss risk decreases
Solution Approach 1:
The patent segments data into domain-specific pools (subscriber data in one pool, billing data in another) rather than creating redundant copies across multiple locations. This segmentation approach maintains reliability through domain isolation while avoiding redundant storage of the same data in multiple places.
Solution Approach 2:
The patent introduces data products as intermediary objects that mediate between data pools and consumers. Data products provide standardized access interfaces to data across domains, enabling reliable data access without requiring redundant storage of the same data in multiple locations.
3Adaptability or versatility
If data is moved between domains, then data accessibility improves, but data loss increases
Solution Approach 1:
The patent uses data products as intermediaries that live in specific data pools and provide standardized access to data. When data needs to be accessed across domains, the data product interface mediates the interaction, maintaining data integrity by keeping data in its original pool while enabling versatile access through standardized protocols.
Solution Approach 2:
The patent replaces mechanical data movement (physical copying and transferring of data between locations) with a virtualized data access model where data remains stationary in its domain pool and consumers interact through standardized data product interfaces, eliminating data loss risks associated with movement.
4Adaptability or versatility
If virtualization and cloud-roaming are implemented, then network flexibility improves, but data management complexity increases
Solution Approach 1:
The patent segments data management into domain-specific pools managed by individual data domain owners, each responsible for their own data products. This segmentation reduces overall system complexity by dividing the complex virtualized network into manageable data domains with clear ownership and responsibility boundaries.
Solution Approach 2:
The patent implements universal data product interfaces that work across all data domains and network scenarios. These standardized interfaces provide multi-functional access capabilities for various network operations (mobility management, billing, analytics) without requiring complex domain-specific data management logic for each scenario.
Data Source
AI summary
A distributed data mesh system for networks is described herein. The distributed data mesh system includes data organized into separate domains, where each domain includes one or more data products representing the data in the domain. The distributed data mesh system additionally includes a data infrastructure catalog that includes information describing the data products and data domains. The distributed data mesh system is also designed based on a set of governing principles which govern how data is created, used, and maintained.