[Back]

M. Ruaro, A. Jantsch, F. Moraes:
"Self-adaptive QoS management of computation and communication resources in many-core SoCs";
ACM Transactions on Embedded Computing Systems, 18 (2019), 4; 1 - 21.

Providing quality of service (QoS) for many-core systems with dynamic application admission is challengingdue to the high amount of resources to manage and the unpredictability of computation and communica-tion events. Related works propose a self-adaptive QoS mechanism concerned either in communication orcomputation resources, lacking, however, a comprehensive QoS management of both. Assuming a many-coresystem with QoS monitoring, runtime circuit-switching establishment, task migration, and a soft real-timetask scheduler, this work fills this gap by proposing a novel self-adaptive QoS management. The contribu-tion of this proposal comes with the following features in the QoS management: (i) comprehensiveness, bycovering communication and computation resources; (ii) online, adopting the ODA (Observe, Decide, Act)runtime closed-loop adaptation; and (iii) reactive and proactive decisions, by using a dynamic applicationprofile extraction technique, which enables the QoS management to be aware of the profile of running appli-cations, allowing it to take proactive decisions based on a prediction analysis. The proposed QoS managementadopts a decentralized organization by partitioning the system in clusters, each one managed by a dedicatedprocessor, making the proposal scalable. Results show that the proactive feature accurately extracts the appli-cations´ profile, and can prevent future QoS violations. The synergy of reactive and proactive decisions wasable to sustain QoS, reducing the deadline miss rate by 99.5% with a severe disturbance in communicationand computation levels, and avoiding deadline misses up to 70% of system utilization.

Computer systems organization, Real-time system architecture, System on a chip, Quality of service, many-core, self-adaptation, prediction

http://dx.doi.org/10.1145/3328755