Papers by Orlando Micolini
2022 IEEE Biennial Congress of Argentina (ARGENCON), Sep 7, 2022
XXVI Congreso Argentino de Ciencias de la Computación (CACIC) (Modalidad virtual, 5 al 9 de octubre de 2020), 2020
The evolution of technology and electronic devices, the widespread use of IoT, and the compliance... more The evolution of technology and electronic devices, the widespread use of IoT, and the compliance with specific regulatory requirements of the industry have made the process of designing embedded systems more complex and challenging. These systems are generally parallel, concurrent, reactive, and/or event-driven. In these systems, the data and events are heterogeneous and non-deterministic as they interact with the external environment. Extended Petri nets constitute an elective platform and system-independent modeling language, which makes it appropriate for modeling embedded systems. To take full advantage of the modeling efforts, it is desirable to use the built models to obtain part of the system implementation. This paper presents the design and implementation of an Extended Petri Processor and its modular architecture. This processor makes use of the extended state equation of Petri Nets, executing the model of the mentioned systems, intending to mitigate the time needed for development, and to reduce programming errors.
En este trabajo, se presenta un procesador de Redes de Petri con Tiempo, el que es la evolucion d... more En este trabajo, se presenta un procesador de Redes de Petri con Tiempo, el que es la evolucion del Procesador de Petri Temporizado. Este procesador es programado directamente con las matrices y vectores del formalismo de Petri, lo que permite aprovechar el poder de las redes de Petri para modelar sistemas de tiempo real y verificar formalmente sus propiedades, evitando errores de programacion al implementar el programa a ejecutar. Este desarrollo ha sido realizado como un IP-cores y es usado en un sistema Multi-core. De esta manera, es posible realizar la implementacion del sistema utilizando este IP-core, lo que asegura las propiedades del modelo realizado con la red de Petri con Tiempo, que verifican los requerimientos del modelo que representa al sistema real, sean cumplido.
Revista Iberoamericana de Tecnología en Educación y Educación en Tecnología, 2013
The execution model proposed by Carlo Ghezzi and Mehdi Jazayeri, known as Simplesem, does not cov... more The execution model proposed by Carlo Ghezzi and Mehdi Jazayeri, known as Simplesem, does not cover languages with parallel execution instructions. The aim is to extend this educational tool model incorporating the existence of multiple processors primitive language that can express basic concepts of parallelism. For this, a tool to analyze graphical and easily the impact of multithreaded programs on low-level instructions that operate directly on the shared memory of a virtual machine was developed. This extension allows to represent the operational semantics of parallel languages that require multi-threaded processors, very common now a days. This tool is intended to be distributed to the educational community, so that studies on the benefits obtained when applied during the process of learning programming languages can be made.
La direccion ha desarrollado en los ultimos anos un proyecto para la generacion de una base de da... more La direccion ha desarrollado en los ultimos anos un proyecto para la generacion de una base de datos con los datos pluviometricos de la provincia, desde 1940 a la actualidad. Para su consulta se diseno una interface tipo web . Dicha base se encuentra cargada en una computadora tipo PC, pero su volumen es tal que el procesamiento requerido para obtener respuestas de consultas particulares o bien de generacion de estadisticas, es inaceptable por lo lento. En virtud de esto se hace necesario el traslado de la base a un entorno de mayor capacidad de procesamiento, que en la actualidad, por una cuestion economica, es conveniente hacer sobre un cluster de computadoras multicore, con capacidad de almacenamiento y procesamiento distribuido. Por lo que el objeto de la transferencia es el conocimiento y procedimiento necesario para trasladar a dicho entorno, el estado actual de la base de datos, y su posterior mantenimiento, puesta a punto y optimizacion del sistema necesario.
Software modularization presents important advantages when it comes to robotic development. That ... more Software modularization presents important advantages when it comes to robotic development. That is based on the definition of well-known limits for each module that make the parallel development easier. Furthermore, that way of working allows the reutilization of one module in more systems. There is a paradigm known as Components Based Software which follows the mentioned idea and needs definitions that include interfaces, responsibilities and certain communication rules between the members of the whole system. Precisely, the JAUS standard defines both a model of components with their interfaces, and an architecture, in the specific context of robotics. There are several SDKs, but they either lack a multi-platform support, or work with platforms whose features do not satisfy the needs of the developers, such as agile development of proof of concepts, robotic specific development resources (existent projects, community support, GUI, among others). This work extends the possibilities...
Este trabajo propone generalizar la ecuación de estado de las redes de Petri, para representar re... more Este trabajo propone generalizar la ecuación de estado de las redes de Petri, para representar redes de Petri no autónomas con distintos tipos de brazos y semánticas temporales. Esta ecuación generalizada facilita la implementación por hardware de un IP-Core para la ejecución de la red. La solución expresada por esta ecuación de estado da origen a un algoritmo que preserva el modelo original, facilita su ejecución en paralelo y permite abordar problemas de mayor tamaño y complejidad. Además, se aborda un caso donde se pone de manifiesto las ventajas de incluir eventos, distintos tipos de brazos y semánticas temporales. Abstract. This work proposes to generalize the state equation of a Petri Net, with the objective of represent different types of arcs and time semantics in non-autonomous Petri nets. This generalized equation facilitates the hardware implementation of an IP-Core to execute the network. The solution that these extended state equation express raises an algorithm that pr...
Reactive and concurrent embedded systems execute restricted algorithms depending on the requireme... more Reactive and concurrent embedded systems execute restricted algorithms depending on the requirements. It is possible to implement one of these hardware-software systems by using a Petri Net Processor. If logic and policy are decoupled from the system actions, then we can improve maintainability and system validation. To achieve this, the Petri Processor is integrated with other traditional processors, forming a heterogeneous multi-core processor, which allows to verify the system using Petri Net mathematical formalisms. In this article, a Modular Petri Processor Architecture is exposed, as well as the inclusion of programmable queues that enhance maintainability, module re-usage and semantic extension.
We call BUS to any group of lines whose main purpose its interconnect different devices. One of t... more We call BUS to any group of lines whose main purpose its interconnect different devices. One of the most critical problem in digital aplicattions its to establish a fast access communication between peripheral devices. With the intention of diminishing the limitations in the data transfer rate, in year 1992, Intel develops the Standard bus of interconnection of peripheral components (PCI). The PCI Bus, is the most used standard for the development of aplication where work togheter external components to the CPU; this paper discuss the development of my end career project, which raises the implementation of a CORE, intrument that allows to fulfill the highest requirements of the communication protocol PCI, this instrument is implemented on a reprogrammable logic FPGA Board. Also, explain the implement of drivers for testing the development under an operating system, and several visual aplications were created to validate the communication and bidirectional data transfer. At the end o...
The evolution of technology and electronic devices, the widespread use of IoT, and the compliance... more The evolution of technology and electronic devices, the widespread use of IoT, and the compliance with specific regulatory requirements of the industry have made the process of designing embedded systems more complex and challenging. These systems are generally multithreaded, parallel, concurrent, reactive, and/or event-driven. In these systems, the data and events are heterogeneous and non-deterministic as they interact with the external environment. Extended Petri nets constitute an elective platform and system-independent modeling language, which makes it appropriate for modeling embedded systems. To take full advantage of the modeling efforts, it is desirable to use the built models to obtain part of the system implementation. This work presents the design and implementation of an Extended Petri Processor, its modular architecture and an algorithm for the automatic determination of the number of active and necessary threads or processes. This processor makes use of the extended state equation of Petri Nets, executing the model of the mentioned systems, aiming to mitigate the time needed for development, and reduce programming errors.
La centralización de la sincronización de procesos mediante un IP-Core disminuye los tiempos requ... more La centralización de la sincronización de procesos mediante un IP-Core disminuye los tiempos requeridos para la labor de sincronización. La implementación de un procesador de Petri que ejecuta el formalismo de redes de Petri brinda una forma sencilla y efectiva de modelar, implementar, ejecutar y programar en forma directa sistemas reactivos con un alto grado de paralelismo. En este artículo se propone el desarrollo de un procesador de redes de Petri Jerárquicas (HPNP) implementado en un IP-Core, el cual, mantiene los beneficios del procesador de Petri, logra una notable reducción en los recursos de hardware utilizados y también la ejecución paralela de los procesos con el procesador de Petri. Esto último es posible gracias a la división de las redes de Petri y la implementación en hardware de un algoritmo que permite la comunicación entre las sub-redes resultantes.
This paper exposes the application of the design and development methodology for an access contro... more This paper exposes the application of the design and development methodology for an access control embedded system, uncoupling the logic, fire policy, and actions. Thus, the proposed methodology makes the final system modular, simple, maintainable, formal, and flexible; allowing for the formal verification of the logic in the earlier stages of development. The system’s logic is modeled using the Petri net´s generalized state equation that enables the resultant logic model to be executable using a monitor. The monitor integrates the Petri net, the policy, and the actions, creating a heterogeneous system that adds the ability to verify the system with the mathematics formalisms of the used model. In addition, the advantages of applying the proposed methodology to the design of an embedded, critical, and reactive real system are revealed in the addressed case. Keywords — Reactive systems design, event-driven architecture, Petri Nets, Concurrent programming, Concurrency Monitor, Design ...
Actually there is a great development interest and use of distributed system. In those systems th... more Actually there is a great development interest and use of distributed system. In those systems the communication network has the most important role in its operation. The complexity of setting up these systems (effort, time, and money) is strongly influenced by different parameters and network configurations. This takes us to propose an instrument for the net simulation of a distributed system. In this paper we present a multi-platform tool, implemented with a parallel architecture, for the evaluation and testing of the parameters in the net for tuning system performance. Our purpose constitutes an integration and extension of existing software for the net simulation. This tool allows the messages capture between subsystems, the injection of these messages into the simulator and its recovery to send them to the target subsystem.
Uploads
Papers by Orlando Micolini