Dynamic and Multidimensional Dataflow Graphs

Abstract

Much of the work to date on data ow models for signal processing system design has focused decidable data ow models that are best suited for onedimensional signal processing. In this chapter, we review more general data ow modeling techniques that are targeted to applications that include multidimensional signal processing and dynamic data ow behavior. As data ow techniques are applied to signal processing systems that are more complex, and demand increasing degrees of agility and exibility, these classes of more general data ow models are of correspondingly increasing interest. We begin with a discussion of two data ow modeling techniques -multi-dimensional synchronous data ow and windowed data ow -that are targeted towards multidimensional signal processing applications. We then provide a motivation for dynamic data ow models of computation, and review a number of speci c methods that have emerged in this class of models. Our coverage of dynamic data ow models in this chapter includes Boolean data ow, the stream-based function model, CAL, parameterized data ow, and enable-invoke data ow.

Modeling pipelined application with Synchronous Data Flow graphs

“…Inclusion of only mapping information indeed would require to force the mutual exclusion between the activation of some actors, but mutual exclusion cannot be modeled in a SDF graph. Boolean SDF graphs overcome this limit, but giving up to statical schedulability [11]. On the contrary, describing a mapping solution given a particular schedule does not require use of mutual exclusions since actors mapped on the same processing element are already forced to be activated in sequence.…”

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

“…Dataflow is widely used as a programming methodology for design and implementation of signal processing systems (e.g., see [3]). In the context of signal processing systems, dataflow provides a formal representation of high level signal flowgraph functionality, which can be leveraged to help verify key implementation properties, and optimize objectives such as performance, memory requirements, and energy consumption.…”

“…Quasi-static scheduling provides useful trade-offs among flexibility, efficiency, and predictability, and is increasingly used in model-based design processes for signal processing [3].…”

“…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.…”

“…Commonly-targeted objectives of optimized schedulers include exploiting parallelism, streamlining buffer management, and minimizing energy or power consumption. For discussions of various scheduling techniques in the context of signal processing system design, we refer the reader to [3].…”