A Practical Framework for Specification, Verification, and Design of Self-Timed Pipelines

Simatic, Jean; Cherkaoui, Abdelkarim; Bertrand, F.; Bastos, Rodrigo Possamai; Fesquet, Laurent

doi:10.1109/async.2017.16

Cited by 10 publications

(1 citation statement)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This will be later useful for interpreting the protocol impact on the circuit behavior when the Timed Petri Net model will be enriched in its SystemVerilog description. The approach can be generalized by considering communication channels, like joins, forks, splits, and merges when the pipeline structure is no longer linear [23]. In our method, the Timed Petri Net model is described in SystemVerilog.…”

Section: Timed Petri Net Modelmentioning

confidence: 99%

High Level Current Modeling for Shaping Electromagnetic Emissions in Micropipeline Circuits

Germain

Engels

Fesquet

2019

JLPEA

Self Cite

View full text Add to dashboard Cite

In order to fit circuit electromagnetic emissions within a spectral mask, a design flow based on high level current modeling for micropipeline circuits is proposed. The model produces a quick and rough estimation of the circuit current, thanks to a Timed Petri Net determining the activation instants of the different micropipeline stages and an asymmetric Laplace distribution modeling the current peaks of the activated stages. The design flow exploits this current estimation for shaping the electromagnetic emissions by setting the controller delays of the micropipeline circuits. The delay adjustment is performed by a genetic algorithm, which iterates until the electromagnetic emissions match the targeted spectral mask. In order to evaluate the technique, an Advanced Encryption Standard (AES) circuit has been designed. We first observed that the obtained current curve fits well with a gate simulation. Then, after shaping the electromagnetic emissions, the simulation shows that the spectrum fits within the spectral mask.

show abstract

Section: Timed Petri Net Modelmentioning

confidence: 99%