Single Address Space Data Processing for Edge Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional IoT approaches for controlling industrial processes face latency and jitter issues due to remote data processing, making it difficult to ensure predictable data processing times in edge computing systems, particularly those using loosely-coupled microservice components across different operating system address spaces.
Innovation Solution
Implementing a data processing system with a communication component using a publish-subscribe messaging pattern within a single memory address space, allowing data processing components to communicate predictably and prioritize data transmission based on topic priority, while optionally using a real-time operating system and multi-core processor affinity techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data processing components are implemented across different operating system address spaces using IPC, then system modularity and flexibility are improved, but data processing predictability and latency control deteriorate
Solution Approach 1:
The patent merges multiple data processing components into a single operating system address space, eliminating the need for inter-process communication (IPC) between them. This consolidation allows direct memory access and eliminates the variability introduced by IPC mechanisms, thereby achieving predictable data processing times while maintaining component functionality through the shared address space.
2Ease of operation
If remote computing devices are used to control embedded systems over the internet, then system accessibility and remote control capability are improved, but data processing latency and jitter increase
Solution Approach 1:
The patent segments the system into an edge computing platform that performs time-sensitive data processing locally, separate from remote control functionality. The edge platform processes critical data within the embedded system's address space with minimal latency, while remote devices communicate with the edge platform for monitoring and non-time-critical control functions, thus resolving the latency issue without sacrificing remote accessibility.
3Ease of manufacture
If conventional operating systems are used in edge computing systems, then ease of deployment and hardware compatibility are improved, but data processing time predictability deteriorates
Solution Approach 1:
The patent changes the fundamental parameter of memory address space management by consolidating all data processing components into a single address space. This parameter change transforms the data processing time from variable (due to IPC overhead in conventional systems) to predictable and controllable, while the system maintains deployment ease through standard operating system frameworks that support shared memory models.
Data Source
AI summary
A data processing system comprising a processor and memory and on which is running an operating system. A computer program executable for implementing a control application for controlling an embedded system is loaded in the memory which controls the operating system to instantiate in an operating system memory a plurality of data processing components and a communication component. The communication component is configured to facilitate data communication between the data processing components using a publish-subscribe messaging pattern. The communication component and data processing components are instantiated in a single memory address space.