Distributed Data Processing Workload Distribution via Edge Terminals

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed data processing systems, optimal resource allocation and data transmission between heterogeneous computers in a cloud network are compromised due to differences in computing power and resource availability, leading to inefficient workload distribution and network traffic.

Innovation Solution

A computer-implemented method that optimizes data processing by transferring workload from servers to end terminals, utilizing separate data exchange components for communication between terminals and servers, and between terminals themselves, with asynchronous protocols to manage authentication, data generation, and inter-terminal connections, ensuring balanced resource loading and efficient data transmission.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a high-performance computer performs the major load of data processing in the computer cloud, then data processing productivity is improved, but network resource consumption for communicating results increases

Engineering Contradiction:
Improvedata processing productivityVSAvoidnetwork resource consumption
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The patent extracts the data processing function from the centralized server and relocates it to edge terminals. The server only stores raw data and receives requests, while actual processing occurs locally on terminals that have downloaded necessary data items, eliminating the need to transmit large volumes of processed results back through the network.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the data processing system into three distinct functional components: the server for data storage and request management, the network for communication, and edge terminals for local data processing. This segmentation allows each component to operate independently and efficiently in its designated role.

Inventive Principle:
Principle #1Segmentation

2Loss of energy

If workload is distributed to end terminals, then network traffic is reduced, but device complexity increases

Engineering Contradiction:
Improvenetwork trafficVSAvoidsystem complexity
Core Design Contradiction:
Loss of energyVSDevice complexity

Solution Approach 1:

The patent implements partial distribution of processing workload to terminals only when necessary. The server maintains the option to process data itself or delegate to terminals based on data characteristics and system state, avoiding the complexity of full distribution while still achieving network traffic reduction for suitable workloads.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The server acts as an intermediary that manages the complexity of coordinating between data storage and distributed processing. It handles request routing, determines which terminals should process which data items, and aggregates results, shielding individual terminals from system-wide complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Speed

If data is stored locally on end terminals, then data access speed is improved, but data protection requirements increase

Engineering Contradiction:
Improvedata access speedVSAvoiddata protection
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The patent applies different quality levels to different data locations. Data stored on terminals is treated with enhanced protection measures compared to centralized storage, reflecting the increased risk of local storage. The system adapts security protocols based on where data resides in the distributed architecture.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3182675B1Transmission of data in a distributed data processing computer system
Publication Date: 2019.02.20 SAP SE
  • EP3182675B1 patent drawingFigure 1
  • EP3182675B1 patent drawingFigure 2
  • EP3182675B1 patent drawingFigure 3~5

AI summary

A computer-implemented method of distributed data processing comprising: establishing an end terminal - server connection between an end terminal and a server; sending a data request from the end terminal to the server via the established end terminal - server connection; terminating the established end terminal - server connection; reestablishing the end terminal - server connection; downloading from the server to the end terminal via the reestablished end terminal -server connection one of the following data items generated by the server in response to receiving the data request: an information data item generated by the server according to the data request or a connectivity data item comprising a connectivity information for establishing a connection to another end terminal, the information data item having a bigger data volume than the connectivity data item; terminating the reestablished end terminal - server connection; in response to downloading from the server of the connectivity data item establishing an end terminal - end terminal connection between the end terminal and the another end terminal using the connectivity information comprised in the connectivity data item; processing the data request, the processing comprising: sending the data request from the end terminal to the another end terminal via the established end terminal - end terminal connection, the end terminal causing the another end terminal to generate an information data item according to the data request via the established end terminal - end terminal connection, and downloading from the another end terminal to the end terminal the information data item generated by the another end terminal via the established end terminal - end terminal connection; and terminating the established end terminal - end terminal connection..