Online Peak Power and Maximum Temperature Management in Multi-core Mixed-Criticality Embedded Systems

Ranjbar, Behnaz; Nguyen, Tuan D. A.; Ejlali, Alireza; Kumar, Akash

doi:10.1109/dsd.2019.00084

Cited by 13 publications

(58 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Socci's approach does not violate TDP constraint because it drops all LC-tasks when the system switches to the HI mode, which means it has zero-percent QoS of LC tasks in the HI mode. On the other hand, methods of [22] and [33] guarantee 90.91% of LC tasks execution in the HI mode, but it frequently violates the TDP constraint (Fig. 7).…”

Section: )mentioning

confidence: 99%

“…However, in this scenario, the TDP constraint is violated several times due to the execution of HC tasks in parallel on cores, and there is no policy to manage the peak power. Moreover, for the second scenario in the methods of [22], [33], the figure shows that in 22.5% of task sets, the TDP constraint is violated while executing some HC tasks on cores. Besides, the QoS in the second scenario for [23] is zero, due to dropping all LC tasks in the HI mode.…”

Section: )mentioning

confidence: 99%

“…Besides, the QoS in the second scenario for [23] is zero, due to dropping all LC tasks in the HI mode. Now, if we investigate the methods of [22], [23], [33] in the third scenario, the QoS of [22], [33] is more than the QoS of [23] due to the executing LC tasks in the HI mode. However, they have less QoS in this scenario compared to our proposed method.…”

Section: )mentioning

confidence: 99%

“…9 shows the average normalized peak power consumption to the TDP constraint and maximum temperature in the worst case that the system switches to the HI mode after executing the first HC task and all tasks execute up to their higher WCET. Hence, in the worst-case scenario, [33] has the same power profile and thermal distribution as [22].…”

Section: This Is An Extended Version Of the Paper Accepted To Be Published In Ieee (Tcad) 2021 Which Includes Appendicesmentioning

confidence: 99%

See 3 more Smart Citations

Toward the Design of Fault-Tolerance-Aware and Peak-Power-Aware Multicore Mixed-Criticality Systems

Ranjbar

Hosseinghorban

Salehi

et al. 2022

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

Self Cite

View full text Add to dashboard Cite

Mixed-Criticality (MC) systems have recently been devised to address the requirements of real-time systems in industrial applications, where the system runs tasks with different criticality levels on a single platform. In some workloads, a highcritically task might overrun and overload the system, or a fault can occur during the execution. However, these systems must be fault-tolerant and guarantee the correct execution of all highcriticality tasks by their deadlines to avoid catastrophic consequences, in any situation. Furthermore, in these MC systems, the peak power consumption of the system may increase, especially in an overload situation and exceed the processor Thermal Design Power (TDP) constraint. This may cause generating heat beyond the cooling capacity, resulting the system stop to avoid excessive heat and halting the processor. In this paper, we propose a technique for dependent dual-criticality tasks in fault-tolerant multi-core MC systems to manage peak power consumption and temperature. The technique develops a tree of possible task mapping and scheduling at design-time to cover all possible scenarios and reduce the low-criticality task drop rate in the highcriticality mode. At run-time, the system exploits the tree to select a proper schedule according to fault occurrences and criticality mode changes. Experimental results show that the average task schedulability is 74.14% on average for the proposed method, while the peak power consumption and maximum temperature are improved by 16.65% and 14.9 • C on average, respectively, compared to a recent work. In addition, for a real-life application, our method reduces the peak power and maximum temperature by up to 20.06% and 5 • C, respectively, compared to a state-ofthe-art approach.

show abstract

Section: )mentioning

confidence: 99%

Section: )mentioning

confidence: 99%

Section: )mentioning

confidence: 99%

Section: This Is An Extended Version Of the Paper Accepted To Be Published In Ieee (Tcad) 2021 Which Includes Appendicesmentioning

confidence: 99%

See 2 more Smart Citations

Toward the Design of Fault-Tolerance-Aware and Peak-Power-Aware Multicore Mixed-Criticality Systems

Ranjbar

Hosseinghorban

Salehi

et al. 2022

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, if, for instance, a failure has happened in High-Criticality Tasks (HCTs) and the execution time of at least one HCT exceeds its Worst-Case Execution Time (WCET), these tasks will overrun and the system switches to the high-criticality (HI) mode. In this situation, if we do not drop or degrade Low-Criticality Tasks (LCTs), the deadline of HCTs (such as safety-critical tasks) may be missed, which can lead to catastrophic consequences [2]- [5], [8], [13]. Nevertheless, the frequent deadline misses or service degradation of some LCTs, such as mission-critical tasks (MCTs), may have a negative impact on the other HCTs and MCTs themselves, and consequently on the entire system and may prevent the system from accomplishing its mission correctly.…”

Section: Introductionmentioning

confidence: 99%

FANTOM: Fault Tolerant Task-Drop Aware Scheduling for Mixed-Criticality Systems

et al. 2020

Self Cite

View full text Add to dashboard Cite

Mixed-Criticality (MC) systems have emerged as an effective solution in various industries, where multiple tasks with various real-time and safety requirements (different levels of criticality) are integrated onto a common hardware platform. In these systems, a fault may occur due to different reasons, e.g., hardware defects, software errors or the arrival of unexpected events. In order to tolerate faults in MC systems, the re-execution technique is typically employed, which may lead to overrun of highcriticality tasks (HCTs), which necessitates the drop of low-criticality tasks (LCTs) or degrading their quality. However, frequent drops or relatively long execution times of LCTs (especially mission-critical tasks) are not always desirable and it may impose a negative impact on the performance, or the functionality of MC systems. In this regard, this paper proposes a realistic MC task model and develops a design-time task-drop aware schedulability analysis based on the Earliest Deadline First with Virtual Deadline (EDF-VD) algorithm. According to this analysis and the proposed scheduling policy based on the new MC task model, in the high-criticality (HI) mode, when an HCT overruns and the system switches to the HI mode, the number of drops per LCT is prohibited from passing a predefined threshold. In addition, to guarantee the real-time constraints and safety requirements of MC tasks in the presence of faults (assuming transient faults in this paper), a corresponding scheduling mechanism has been developed. According to the obtained results from an extensive set of simulations, which have been validated through a realistic avionic application, the proposed method improves the acceptance ratio by up to 43.9% compared to state-of-the-art.

show abstract