Software synthesis from the dataflow interchange format

Abstract

Specification, validation, and synthesis are important aspects of embedded systems design. The use of dataflow-based design environments for these purposes is becoming increasingly popular in the domain of digital signal processing (DSP). The dataflow inter-change format (DIF) [11] and the associated DIF package have been developed for specoeying, working with, and transferring dataflow-based DSP designs across tools. In this paper, we present the newly developed DIF-to-C software synthesis framework for automatically generating monolithic C-code implementations from DSP system specifications that are programmed in DIF. This framework allows designers to efficiently explore the complex range of implementation tradeofJ~ that are available through various dataflow-based techniques for scheduling and memory management. Furthermore, the DIF-to-C framework provides a standard, vendor-neutral mechanism for linking coarse grain data-flow optimizations with fine grain handoptimized libraries and the large body of optimization techniques in the area of C compilers for DSP. 7Through experiments involving several DSP applications, we demonstrate the novel and useful capabilities of our DIF-to-C software synthesis framework.

Rapid prototyping for digital signal processing systems using Parameterized Synchronous Dataflow graphs

“…We plan to release PSDFsim as part of the DIF project [10,15], which provides open-access tools -primarily in the form of Java-based jar files -that can be used for compiling, analyzing, and extending specifications in the dataflow interchange format (the DIF language).…”

“…Functional DIF allows actors whose internals are programmed in Java to be integrated with EIDF-based dataflow graphs that are specified in the dataflow interchange format (DIF). DIF is a textual language for specifying dataflow graphs in terms of arbitrary dataflow models of computation [10].…”

Modeling and optimization of dynamic signal processing in resource-aware sensor networks

“…This form of token transfer predictability promotes the development of powerful techniques for analysis and optimization, and there is a large and growing body of literature on such SDF techniques (e.g., see [13,8,5,16,7,25,3]. However, the restriction to constant token transfer rates significantly limits the expressive power of SDF.…”

“…A major objective in the dataflow interchange format (DIF) project is the development of efficient methods for design and implementation of signal processing systems from static and dynamic dataflow representations [8]. In this paper, we provide a discussion of different dataflow models of computation that are being developed in DIF project to help address the challenges of structured specification, simulation, and synthesis of dynamic signal processing systems.…”

Exploiting Statically Schedulable Regions in Dataflow Programs

“…The semantics that underlies the CAL language bears some similarity to the stream-based functions model of computation [6]. DIF [2] is a language for specifying dataflow graphs in terms of subsystems that conform to different kinds of specialized dataflow modeling techniques, and The DIF Package (TDP) is a tool for analyzing DIF language specifications, with emphasis on scheduling-and memory-managementrelated analysis techniques [2]. CAL2C [3] is a tool that performs automatic generation of C code from CAL networks, thereby providing a direct bridge between CAL and off-the-shelf embedded processing platforms.…”