SMAP Software-Hardware Partitioning via SPLIT Architecture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining the optimal software-hardware partitioning for Scheduled Medium-Access Protocols (SMAPs) in communication systems is challenging, as existing methods lack a simple and efficient approach that allows complex application-dependent code to be implemented in software.
Innovation Solution
The Scheduling Protocol Lightweight Integration Testbed (SPLIT) architecture partitions SMAP functionality between software and hardware, with a scheduler implemented in software and a grant generator in hardware, enabling independent operation and improving system performance by decoupling scheduling from grant generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If SMAP functionality is implemented entirely in software, then complex application-dependent code can be implemented flexibly, but system performance and grant-generation efficiency deteriorate
Solution Approach 1:
The patent divides the SMAP functionality into two separate modules: a scheduler implemented in software that handles complex application-dependent logic, and a grant generator implemented in hardware that handles time-critical grant generation. This segmentation allows each module to be optimized for its specific requirements, resolving the contradiction between software flexibility and hardware efficiency.
2Productivity
If SMAP functionality is implemented entirely in hardware, then grant-generation efficiency is improved, but adaptability to different protocols and complex code implementation deteriorates
Solution Approach 1:
The patent separates protocol-adaptive functions (scheduler) from time-critical functions (grant generator). The software scheduler can be modified to support different SMAP variants, while the hardware grant generator maintains high-speed operation. This segmentation resolves the contradiction between hardware efficiency and protocol adaptability.
Solution Approach 2:
The patent introduces an intermediary interface between the software scheduler and hardware grant generator. This interface allows the software to dynamically control the hardware grant generator without requiring direct hardware modification, enabling protocol adaptability while maintaining hardware efficiency.
3Device complexity
If scheduling and grant generation are coupled, then system complexity is reduced, but scheduler quality and occupancy become dependent on grant-generation rate
Solution Approach 1:
The patent decouples the scheduler and grant generator into independent modules with well-defined interfaces. The scheduler operates independently to determine optimal grant allocations based on traffic conditions, while the grant generator independently executes grant transmission. This decoupling resolves the contradiction by allowing scheduler quality to be evaluated independently of grant-generation rate, while maintaining manageable system complexity through clear module boundaries.
Data Source
AI summary
A processing device, configured to implement at least a portion of a scheduled medium-access protocol (SMAP) in a communication system, comprises a processor, a memory coupled to the processor, and one or more additional hardware modules. The functionality of the portion of the SMAP implemented in the processing device is partitioned between software, stored in the memory and executable by the processor, and hardware comprising the one or more additional hardware modules. In an illustrative embodiment, the processing device comprises a head-end device of a passive optical network, and the functionality comprises at least a scheduler and a grant generator, with the scheduler being implemented in the software and the grant generator being implemented in the hardware.