ASTROLABE: A Rigorous Approach for System-Level Performance Modeling and Analysis

Ayoub Nouri (1) , Marius Bozga (1) , Anca Molnos (2) , Axel Legay (3) , Saddek Bensalem (1)

1 VERIMAG - IMAG - VERIMAG
2 CEA-LETI - Commissariat à l'énergie atomique et aux énergies alternatives - Laboratoire d'Electronique et de Technologie de l'Information
3 TAMIS - Threat Analysis and Mitigation for Information Security

Building abstract system-level models that faithfully capture performance and functional behavior for embedded systems design is challenging. Unlike functional aspects, performance details are rarely available during the early design phases, and no clear method is known to characterize them. Moreover, once such models are built, they are inherently complex as they mix software models, hardware constraints, and environment abstractions. Their analysis by using traditional performance evaluation methods is reaching the limit. In this article, we present a systematic approach for building stochastic abstract performance models using statistical inference and model calibration, and we propose statistical model checking as a scalable performance evaluation technique for them.

Mots clés: C4 [Performance of Systems]: Software/Program Verification; I64 [Model Validation and Analysis]; I65 [Model Development]; G3 [Probability and Statistics] General Terms: Design; Performance; Stochastic Models; Statistics; Formal Methods Additional Key Words and Phrases: Many-cores; performance evaluation; model-based design

Domaines: Informatique [cs]

Ayoub Nouri, Marius Bozga, Anca Molnos, Axel Legay, Saddek Bensalem. ASTROLABE: A Rigorous Approach for System-Level Performance Modeling and Analysis. ACM Transactions on Embedded Computing Systems (TECS), 2016, ⟨10.1145/2885498⟩. ⟨hal-01406474⟩