دانلود مقاله ISI انگلیسی شماره 9296
عنوان فارسی مقاله

مدل شبیه سازی گسسته - پیوسته عمومی برای اعتبار سنجی دقیق در طراحی سیستم های ناهمگون

کد مقاله سال انتشار مقاله انگلیسی ترجمه فارسی تعداد کلمات
9296 2007 11 صفحه PDF سفارش دهید محاسبه نشده
خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.
عنوان انگلیسی
Generic discrete–continuous simulation model for accurate validation in heterogeneous systems design
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Microelectronics Journal, Volume 38, Issues 6–7, June–July 2007, Pages 805–815

کلمات کلیدی
/ - مدل شبیه سازی پیوسته - گسسته - مدل سنکرون - لینک های شبیه سازی در نرم افزار مطلب - سیستم
پیش نمایش مقاله
پیش نمایش مقاله مدل شبیه سازی گسسته - پیوسته عمومی برای اعتبار سنجی دقیق در طراحی سیستم های ناهمگون

چکیده انگلیسی

Heterogeneous systems combining several technologies will potentially overcome several performances and applicability limitations, as well as providing new useful features. However, their design is currently confronted to important challenges. One of the key challenges is the integration of pre-built components specific to different application domains (e.g. electrical, mechanical, optical, etc.). In this context, new CAD tools are mandatory to offer a global view of the designed systems and to enable their overall validation. This paper presents a generic discrete–continuous simulation model for an accurate global validation in heterogeneous systems design. This model enables to use powerful tools for the discrete and continuous domains. Solutions are proposed for model's layers implementation in the case of SystemC and Simulink simulators. The continuous–discrete simulation tool is proposed. This tool integrates the two previous simulators by automatically generating global simulation model instances. The evaluation of the simulation model was performed using an illustrative application.

مقدمه انگلیسی

Modern systems like mixed-signal systems and real-time controllers integrate discrete and continuous components. Nowadays, systems on chip are the main drivers for convergence of multiple technologies. In 2001, FRAM and FPGA were integrated on chip then follow in 2003 MEMS and chemical sensors; in 2005 electro-optical; and in 2006 electro-biological. The 2006 ITRS update edition [1] highlights that “the development of new technologies that are placed side-by-side to a digital design in a silicon die presents a whole set of new challenges. Examples are MEMS, electro-optical devices and electro-biological devices. These new components will require modeling of both the interface between the digital portion and the non-digital components and a proper abstraction of the non-digital system behavior in order to still be able to verify the digital portion of the system”. These systems will be ubiquitous in communications, automotive, medical and other domains. The global validation of these systems requires new techniques enabling reusability, high abstraction levels and simulation accuracy from a time point of view. Currently, in order to respect the tight constraints of time-to-market, cost and performance, the heterogeneous systems are designed by reusing pre-designed components. Thus the design flow may be characterized by two main aspects: (1) building components that may be reused and (2) helping in their integration. This type of design represents an important challenge; one of the key issues being the integration of the pre-built components specific to different application domains (e.g. electrical, mechanical, optical, etc.) [2] and [3]. Nowadays, designers build different components to be integrated by using powerful existing tools specialized for a specific application domain (e.g. SystemC for the electronic digital part, Matlab/Simulink for the mechanical part, etc.) and they often prefer to keep using their current tools. Consequently, new CAD tools for heterogeneous systems must be based on global simulation models defined independently from specific languages or simulators. This permits the integration of best and adequate existing tools in order to exploit their full capabilities. To reach this requirement, the co-simulation technique is used in this work, even though it decreases the simulation speed, it allows the designer to describe each, the continuous and the discrete model, in a specific and appropriate language. The co-simulation model must be generic (independent from languages or simulators) in order to respect the mentioned requirement. Some difficulties in the definition of a continuous/discrete co-simulation model are (1) the heterogeneity of the definitions of the concepts manipulated by the discrete and the continuous components and (2) the need of continuous/discrete communication and synchronization. This is eased by using co-simulation interfaces. These interfaces have a great influence on the accuracy and the performance of the global simulation. Their automatic generation is very important, since their design is time consuming and an important source of errors.

نتیجه گیری انگلیسی

Nowadays, in order to cope with complexity and to respect time-to-market constraints, the design of heterogeneous systems is performed by assembling pre-designed components specific to different application domains. This methodology requires new CAD tools for the global specification and validation. This paper presented a generic simulation model providing enough semantics for an accurate global validation of heterogeneous systems integrating continuous and discrete models. The simulation model was implemented by simulation interfaces, generated automatically by the CODIS tool in order to produce global simulation model instances in the case of SystemC and Simulink simulators. The experiments have shown simulation accuracy and an overhead of 25%. As perspectives, we plan (1) to implement the simulation model for Simulink/ModelSim co-simulation and (2) to improve the simulation speed by minimizing simulators interactions and optimizing the IPC-used methods.

خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.