Citation
Alessandro Pinto. "Communication Synthesis with Applications to On-Chip Communication and Building Automation Systems". Talk or presentation, 27, November, 2007.

Abstract

Integration is the next big challenge in Embedded System Design (ESD). The complexity of electronics, from chips to building automation systems, increases at each product generation, while time-to-market shrinks. To overcome the possible barriers coming from a linear increase of productivity, it has become common to assemble systems out of off-the shelf components. This approach to ESD brings up many issues that are not related to the components themselves but rather to the way in which they interact. The problem of interconnecting components correctly appears at all levels of abstraction. At the design entry level, each piece of the system is usually described with the model of computation that most naturally captures its behavior. When the entire system is put together, the interaction between different parts requires the definition of the interaction among heterogeneous models. At lower levels of abstraction, the problems reduces to the design of the actual communication architecture that allows the components to communicate. Correctness at this level is defined also with respect to performance metrics like latency, bandwidth and reliability.

Thus, given a collection of agents and the way in which they communicate (usually specified by a set of point-to-point communication requirements) the embedded system designer faces the problem of defining a communication structure such that the communication requirements are satisfied and the cost is minimized. This problem is not trivial considering the number of different ways in which the abstract point-to-point specification can be refined into a concrete communication structure. However, the flexibility provided by the plethora of communication technologies available today, is a great opportunity to “guarantee performance at minimum cost”: a motto for the embedded system community.

I will present our on-going research effort toward the development of an infrastructure for networked embedded systems. I will motivate our work using two design drivers: on-chip communication and building automation systems, and I will argue that the same theory can be applied independently from the specific application. Then, I will show how the theory has been used to develop a tool for the automatic design of communication structures and I will present some preliminary results. I will conclude with a list of research directions.

Electronic downloads

• pinto_diss_talk.zip · application/zip · 14778 kbytes

• HTML

  Alessandro Pinto. <a
  href="http://chess.eecs.berkeley.edu/pubs/378.html"><i>Communication
  Synthesis with Applications to On-Chip Communication and
  Building Automation Systems</i></a>, Talk or
  presentation,  27, November, 2007.

• Plain text

  Alessandro Pinto. "Communication Synthesis with Applications
  to On-Chip Communication and Building Automation Systems".
  Talk or presentation,  27, November, 2007.

• BibTeX

  @presentation{Pinto07_CommunicationSynthesisWithApplicationsToOnChipCommunication,
      author = {Alessandro Pinto},
      title = {Communication Synthesis with Applications to
                On-Chip Communication and Building Automation
                Systems},
      day = {27},
      month = {November},
      year = {2007},
      abstract = {
  Integration is the next big challenge in
                Embedded System Design (ESD). The complexity of
                electronics, from chips to building automation
                systems, increases at each product generation,
                while time-to-market shrinks. To overcome the
                possible barriers coming from a linear increase of
                productivity, it has become common to assemble
                systems out of off-the shelf components. This
                approach to ESD brings up many issues that are not
                related to the components themselves but rather to
                the way in which they interact. The problem of
                interconnecting components correctly appears at
                all levels of abstraction. At the design entry
                level, each piece of the system is usually
                described with the model of computation that most
                naturally captures its behavior. When the entire
                system is put together, the interaction between
                different parts requires the definition of the
                interaction among heterogeneous models. At lower
                levels of abstraction, the problems reduces to the
                design of the actual communication architecture
                that allows the components to communicate.
                Correctness at this level is defined also with
                respect to performance metrics like latency,
                bandwidth and reliability.
   
   Thus, given a
                collection of agents and the way in which they
                communicate (usually specified by a set of
                point-to-point communication requirements) the
                embedded system designer faces the problem of
                defining a communication structure such that the
                communication requirements are satisfied and the
                cost is minimized. This problem is not trivial
                considering the number of different ways in which
                the abstract point-to-point specification can be
                refined into a concrete communication structure.
                However, the flexibility provided by the plethora
                of communication technologies available today, is
                a great opportunity to “guarantee performance at
                minimum cost”: a motto for the embedded system
                community.
   
   I will present our on-going
                research effort toward the development of an
                infrastructure for networked embedded systems. I
                will motivate our work using two design drivers:
                on-chip communication and building automation
                systems, and I will argue that the same theory can
                be applied independently from the specific
                application. Then, I will show how the theory has
                been used to develop a tool for the automatic
                design of communication structures and I will
                present some preliminary results. I will conclude
                with a list of research directions.
  },
      URL = {http://chess.eecs.berkeley.edu/pubs/378.html}
  }
  

Posted by Douglas Densmore on 28 Nov 2007.
For additional information, see the Publications FAQ or contact webmaster at chess eecs berkeley edu..

Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright.