Prime Scheduling for Computing Task Coincidence Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern computing systems, especially in-vehicle systems, face overload issues due to computing task coincidences, where multiple periodic tasks coincide, leading to increased workload, time delays, and resource consumption, which are typically addressed by reducing tasks or adding resources, both of which are costly or reduce functionality.
Innovation Solution
Implementing prime scheduling by assigning time periods to periodic computing tasks based on prime numbers, reducing the likelihood of coincidences by increasing the hyperperiod, thereby minimizing overload without sacrificing functionality or increasing costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of periodic computing tasks is reduced, then system overload is reduced, but system functionality is reduced
Solution Approach 1:
The patent changes the time period parameter of periodic computing tasks from conventional integer multiples of the common tick to prime number multiples. This parameter transformation fundamentally alters the coincidence pattern of tasks, distributing them more evenly across ticks and preventing overload while preserving all tasks and their functionality.
2Reliability
If start-time offsets are introduced to reduce task coincidences, then system overload is reduced, but implementation complexity increases due to lack of support in automotive software
Solution Approach 1:
Instead of modifying task start times through offsets, the patent changes the fundamental parameter of task period assignment to prime number multiples. This approach achieves load distribution without requiring support for start-time offsets in automotive software, thereby avoiding increased implementation complexity.
3Reliability
If more computational resources are made available, then system overload is reduced, but system cost increases
Solution Approach 1:
The patent addresses overload by changing the temporal distribution parameter of task execution through prime number scheduling, rather than increasing computational resources. This achieves better load balancing and prevents overload while maintaining the same hardware and resource infrastructure, avoiding increased system cost.
4Ease of operation
If periodic computing tasks are scheduled at integer multiples of common tick, then scheduling simplicity is maintained, but task coincidences increase leading to overload
Solution Approach 1:
The patent modifies the scheduling parameter from integer multiples of the common tick to prime number multiples. This change maintains scheduling simplicity by still using regular intervals, while the prime number property fundamentally reduces task coincidences and prevents overload.
Data Source
AI summary
Systems and techniques that facilitate computing task coincidence reduction via prime scheduling are provided. In various embodiments, a system can comprise an input component that can access a periodic computing task performed by a computer system. In various instances, the system can further comprise a prime component that can assign a time period to the periodic computing task. In various cases, a duration of the time period can be a product of a common tick and a prime number. In various embodiments, the system can comprise an execution component that can periodically execute the periodic computing task according to the time period.