The large size and complexity of the modern embedded systems pose great challenges to design and validation. At the so called electronic system level (ESL), designers start with a specification model of the system and follow a systematic top-down design approach to refine the model to lower abstraction levels step-by-step by adding implementation details. ESL models are usually written in C-based System-level Description Languages (SLDLs), and contain the essential features, such as clear structure and hierarchy, separate computation and communication, and explicit parallelism. The validation of ESL models typically relies on simulation. Fast yet accurate simulation is highly desirable for efficient and effective system design.
In this dissertation, we present out-of-order parallel discrete event simulation (OoO PDES), a novel approach for efficient validation of system-level designs by exploiting the parallel capabilities of todays multi-core PCs for system level description languages. OoO PDES breaks the global simulation-cycle barrier of traditional DE simulation by localizing the simulation time into each thread, carefully delivering notified events, and handling a dynamic management of simulation sets. Potential conflicts caused by parallel accesses to shared variables and out-of-order thread scheduling are prevented by an advanced predictive static model analyzer in the compiler. As such, OoO PDES allows the simulator to effectively exploit the parallel processing capability of the multi-core system to achieve fast speed simulation without loss of simulation and timing accuracy.
We perform simulation experiments on both highly parallel benchmark examples and real-world embedded applications, including a JPEG image encoder, an edge detector, a MP3 audio decoder, a H.264 video decoder, and a H.264 video encoder. Experimental results show that our approach can achieve significant simulation speedup on multi-core simulation hosts with negligible compilation cost.
Based on our parallel simulation infrastructure, we then propose a tool flow for dynamic race condition detection to increase the observability for parallel ESL model development. This helps the designer to quickly narrow down the debugging targets in faulty ESL models with parallelism. This approach helps to reveal a number of risky race conditions in our in- house embedded multi-media application models and enabled us to safely eliminate these hazards. Our experimental results also show very little overhead for race condition diagnosis during compilation and simulation.
Overall, our work provides an advanced parallel simulation infrastructure for efficient and effective system-level model validation and development. It helps embedded system designers to build better products in shorter time.
|Commitee:||Demsky, Brian, Gajski, Daniel D.|
|School:||University of California, Irvine|
|Department:||Electrical and Computer Engineering|
|School Location:||United States -- California|
|Source:||DAI-B 75/01(E), Dissertation Abstracts International|
|Subjects:||Computer Engineering, Electrical engineering, Computer science|
|Keywords:||Embedded computer systems, Parallel discrete event simulation, Segment graph, Static code analysis, System-level description language, System-level design|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be