Multi-dimensional Index Hyper-sphere Data Distribution

Solution Overview

Problem

Distributed data storage systems face inefficiencies in processing resources and storage capacity due to the limitations of state-less algorithms when handling data sets with multiple defining fields (keys), leading to scalability issues and increased storage requirements for indexing, which complicates data location and reallocation.

Innovation Solution

The implementation of a data distribution and location method using a metric space with hyper-spheres, where data items are mapped to points in a multi-dimensional space, allowing for efficient data retrieval and storage by calculating distances to hyper-sphere centers and dynamically updating hyper-sphere centers based on historical data to maintain balanced distribution across computers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If state-less algorithms are used for data location, then processing efficiency is improved and storage requirements are reduced, but data re-location becomes difficult and flexibility is lost

Engineering Contradiction:

Improveprocessing efficiencyVSAvoiddata re-location flexibility

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The system pre-calculates and stores hyper-sphere centers and radii during data insertion, creating a spatial index structure in advance. This preliminary action enables efficient O(1) retrieval by comparing query points against pre-computed hyper-spheres, avoiding the need for linear scans while maintaining the ability to relocate data by simply updating hyper-sphere parameters rather than physical positions

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent transforms the traditional flat data storage model into a multi-dimensional metric space using hyper-spheres. Data items are represented as points in this space, and hyper-spheres define regions containing multiple data points. This dimensional transformation enables efficient spatial queries and provides flexibility for data re-location by adjusting hyper-sphere centers and radii without requiring linear search or complex index restructuring

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If state-full algorithms are used for data location, then data re-location flexibility is improved, but processing bottlenecks occur and storage space increases

Engineering Contradiction:

Improvedata re-location flexibilityVSAvoidprocessing speed

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent extracts the essential spatial information (hyper-sphere centers and radii) from the complete data set and stores only this condensed representation. Instead of storing information about every data item's location, the system stores minimal hyper-sphere parameters that define regions containing multiple data items. This extraction reduces storage requirements while maintaining re-location flexibility through hyper-sphere parameter updates

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system pre-computes hyper-sphere centers and radii during data insertion, creating a spatial index structure in advance. This preliminary action enables efficient O(1) retrieval by comparing query points against pre-computed hyper-spheres, avoiding the need for linear scans while maintaining the ability to relocate data by simply updating hyper-sphere parameters rather than physical positions

Inventive Principle:

Principle #10Preliminary action

3Device complexity

If linear scan approach is used for data retrieval, then implementation simplicity is improved, but processing resources grow linearly with data size

Engineering Contradiction:

Improvealgorithm simplicityVSAvoidprocessing resource efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent transforms the traditional flat data storage model into a multi-dimensional metric space using hyper-spheres. Data items are represented as points in this space, and hyper-spheres define regions containing multiple data points. This dimensional transformation enables efficient spatial queries and provides flexibility for data re-location by adjusting hyper-sphere centers and radii without requiring linear search or complex index restructuring

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The system pre-calculates and stores hyper-sphere centers and radii during data insertion, creating a spatial index structure in advance. This preliminary action enables efficient O(1) retrieval by comparing query points against pre-computed hyper-spheres, avoiding the need for linear scans while maintaining the ability to relocate data by simply updating hyper-sphere parameters rather than physical positions

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9292559B2Data distribution/retrieval using multi-dimensional index

Publication Date: 2016.03.22 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

A distributed data storage system uses a data distribution and location algorithm based on distance functions and hyper-spheres in a multi-dimensional space. The distributed data storage system uses the algorithm to maintain, over time, a balanced distribution across a number of computers interconnected by a network of a varying set of data items. Each data item includes one or more key fields. The system also includes an efficient partial-match and exact-match search across a whole set of data items using as search criteria the values of any or all of the sought data item's key field(s).