TTE Network Scheduling via Demand Weighted Bin Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current algorithms for scheduling in Time-Triggered Ethernet (TTE) networks fail to converge on a valid global schedule for proper traffic throughput due to the complexity of data sets with low-rate, low-latency data, which violates assumptions about rate and latency requirements.
Innovation Solution
A method and system for determining weights proportional to the bandwidth and latency demand of each Virtual Link (VL) in the network, using harmonic bin lengths and green zones to sort and schedule VLs, incorporating machine learning for dynamic scheduling and bin selection optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If current scheduling algorithms are used to sort VLs based on transmit rate and latency constraints, then high-rate traffic can be scheduled effectively, but low-rate, low-latency traffic fails to converge on a valid global schedule
Solution Approach 1:
The patent transforms the scheduling problem by changing the parameter used for sorting VLs from transmit rate to demand weight. The demand weight is calculated as a function of both transmit rate and green zone size (w = rate × green_zone), which captures the true scheduling difficulty. This parameter transformation allows the algorithm to handle low-rate, low-latency traffic effectively while maintaining performance for high-rate traffic.
2Manufacturing precision
If tight latency constraints are imposed on low-rate VLs, then latency requirements are met, but the scheduling algorithm fails to converge on a valid global schedule
Solution Approach 1:
The patent performs preliminary calculation of demand weights for all VLs before the actual scheduling process. By pre-computing the demand weight (w = rate × green_zone) for each VL, the algorithm prepares the sorting criterion in advance, enabling it to handle tight latency constraints without failing to converge. This preliminary action ensures that even VLs with strict latency requirements can be properly prioritized.
3Ease of manufacture
If traditional bin selection methods are used, then simple scheduling is achieved, but bin packing efficiency is suboptimal for diverse traffic demands
Solution Approach 1:
The patent applies local quality by treating different VLs with different sorting criteria based on their individual characteristics. Instead of using a uniform sorting approach, the demand weight calculation (w = rate × green_zone) assigns different priorities to different VLs according to their specific rate and latency requirements. This localized approach optimizes bin packing efficiency for each VL's particular needs while maintaining overall scheduling simplicity.
Data Source
AI summary
Methods, systems and apparatuses for scheduling a plurality of Virtual Links (VLs) in a Time-Triggered Ethernet (TTE) network by determining a weight for each of the plurality of VLs, the determined weight being proportional to a demand each of the plurality of VLs will place on the time-triggered ethernet network; generating a plurality of bins whose length in time is harmonic to all the scheduled rates of the plurality of VLs; determining a demand value proportional to how often the bin is expected to be used based upon a green zone of each of the plurality of VLs and the determined weight for each of the plurality of VLs; updating the demand value for each bin within the green zone; sorting the plurality of bins from least demanded to most demanded based upon the updated demand value; and scheduling the sorted plurality of VLs within the sorted plurality of bins.