Transaction-level modeling (TLM) is nowadays a promising design style to deal with the increasing complexity of modern embedded systems. A TLM-based flow involves different verification techniques according to the system being designed and the exhaustiveness of the results to be achieved. The design flow may involve a transition from or to a lower abstraction level (RTL), both for design and verification purposes. Refinement from TLM to RTL is performed to move closer to the physical realization of the system being designed. On the other hand, abstraction from RTL to TLM allows to reuse third-party or already developed components and to integrate them into a system-level design, thus gaining a reduction of design time and costs and an increase in simulation speed. Finally, a further refinement step consists of gate-level synthesis, which brings the design even closer to its physical realization. At this level, test generation is typically a time-consuming activity, so techniques can be adopted to reduce such computation times.

Design and Verification Techniques for TLM-based Design Flows

GUARNIERI, Valerio
2013

Abstract

Transaction-level modeling (TLM) is nowadays a promising design style to deal with the increasing complexity of modern embedded systems. A TLM-based flow involves different verification techniques according to the system being designed and the exhaustiveness of the results to be achieved. The design flow may involve a transition from or to a lower abstraction level (RTL), both for design and verification purposes. Refinement from TLM to RTL is performed to move closer to the physical realization of the system being designed. On the other hand, abstraction from RTL to TLM allows to reuse third-party or already developed components and to integrate them into a system-level design, thus gaining a reduction of design time and costs and an increase in simulation speed. Finally, a further refinement step consists of gate-level synthesis, which brings the design even closer to its physical realization. At this level, test generation is typically a time-consuming activity, so techniques can be adopted to reduce such computation times.
2013
Inglese
TLM; RTL-to-TLM abstraction; TLM-to-RTL refinement; Fault simulation; Fault injection; Fault simulation acceleration
164
File in questo prodotto:
File Dimensione Formato  
guarnieri_phd_thesis.pdf

accesso solo da BNCF e BNCR

Dimensione 3.87 MB
Formato Adobe PDF
3.87 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/182885
Il codice NBN di questa tesi è URN:NBN:IT:UNIVR-182885