Repository logo
 
No Thumbnail Available
Publication

Shared Resource Contention Aware Schedulability Analysis for Multiprocessor Real-Time Systems

Use this identifier to reference this record.
Name:Description:Size:Format: 
COM_CISTER-TR-221202_2023.pdf263.85 KBAdobe PDF Download

Advisor(s)

Abstract(s)

Multicore platforms share the hardware resources such as caches, interconnects, and main memory among all the cores. Due to such sharing, tasks running on different cores compete to access these shared resources which can potentially result in shared resource contention. This shared resource contention can increase the execution times of tasks in a non-deterministic manner. Consequently, the shared resource contention is problematic for hard real-time systems, i.e., systems that run tasks with stringent timing requirements. To address this issue, this PhD dissertation builds novel solutions to model and analyze the shared resource contention that can be suffered by tasks executing on a multicore system. The shared resource contention aware schedulability analysis is then derived by integrating the maximum shared resource contention that can be suffered by the tasks.

Description

This PhD dissertation has resulted in six publications in which one of the papers received Best Paper Award at ICESS 2021. All the papers were published in reputed venues for real-time systems research, i.e., RTSS 2020, RTNS 2021, ICESS 2021, RTCSA 2022, RTSS 2022, Elsevier’s Journal of System Architecture. Two more papers are expected to be published soon.

Keywords

Citation

Organizational Units

Journal Issue