Hardware-Accelerated TSN Configuration for Large-Scale Stream Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing TSN configuration solutions fail to provide efficient, deterministic, and rapid scheduling for large-scale networks with complex data streams, leading to significant time delays and inefficiencies in network deployment, and do not effectively leverage hardware platforms' properties for optimized scheduling.
Innovation Solution
A method and system for efficient TSN configuration based on hardware acceleration, utilizing a scheduling engine that includes hyperperiod and conflict-group based parallel scheduling, which acquires, preprocesses, and schedules data stream attributes to generate and deploy configuration schemes on FPGA platforms, optimizing scheduling complexity and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If centralized configuration architecture with CUC and CNC units is used to configure TSN parameters, then deterministic transmission can be achieved, but configuration complexity and deployment time increase significantly for large-scale networks
Solution Approach 1:
The patent segments the centralized configuration architecture into modular functional units including CUC (Centralized User Configuration), CNC (Centralized Network Configuration), and CNA (Centralized Network Administration) units. Each unit handles specific configuration tasks independently, allowing parallel processing and reducing overall configuration complexity for large-scale TSN networks while maintaining deterministic transmission guarantees.
2Reliability
If SMT and ILP modeling approaches are used for scheduling TS streams, then deterministic and real-time transmission is ensured, but scheduling runtime becomes too long for large-scale networks
Solution Approach 1:
The patent implements preliminary action by pre-calculating and storing scheduling parameters, resource allocation patterns, and transmission timing information in configuration databases before actual TSN operation. The CNC unit generates scheduling schemes in advance using optimized algorithms, and these pre-computed configurations are then rapidly deployed to network devices, significantly reducing real-time scheduling runtime while maintaining deterministic transmission guarantees.
3Productivity
If incremental scheduling and group scheduling are implemented to balance schedulability and runtime, then some improvement is achieved, but scheduling requirements for large-scale networks with massive data streams are still not met
Solution Approach 1:
The patent transitions from traditional single-dimension sequential scheduling to multi-dimensional parallel scheduling by implementing hyperperiod-based parallel scheduling mechanisms. The system divides the scheduling problem into multiple independent time slots and parallel processing streams, allowing simultaneous calculation of scheduling parameters for multiple data streams across different dimensions (time, space, priority levels). This dimensional expansion enables the system to handle massive data streams in large-scale networks while maintaining both high schedulability and efficient runtime performance.
4Ease of operation
If serialized time slot allocation process is used in FITS method, then GCL design can be achieved, but computing time increases significantly in large-scale network environments
Solution Approach 1:
The patent applies copying by creating and utilizing multiple parallel copies of the time slot allocation process simultaneously. Instead of sequentially allocating time slots one after another, the system generates multiple identical allocation computation instances that operate in parallel, each handling different portions of the GCL configuration. This parallel copying approach maintains the ease of GCL design through standardized allocation patterns while dramatically reducing total computing time by eliminating sequential bottlenecks in large-scale network environments.
Data Source
AI summary
The present invention discloses a method and system for efficient configuration of a time-sensitive network based on hardware acceleration, pertaining to the field of wired communication networking technology. The method includes an acquisition step, a preprocessing step, a scheduling step, a generation step and a deployment step. The scheduling step, according to an operating state of a scheduling engine, based on attribute information of to-be-configured data streams, selects a time slot length, calculates a hyperperiod, depending on a size of the hyperperiod, carries out any of hyperperiod based parallel scheduling or conflict-group based parallel scheduling, obtaining a scheduling result, and thereby generates and deploys a configuration scheme to network devices. The present invention is able to, based on on-site network and terminal device conditions, derive global deterministic scheduling and configuration schemes and enables rapid distribution and deployment.