Advanced

Exploiting statically schedulable regions in dataflow programs

Gu, Ruirui; Janneck, Jörn LU ; Raulet, Mickaël and Bhattacharyya, Shuvra S. (2011) In Journal of Signal Processing Systems 63(1). p.129-142
Abstract (Swedish)
Abstract in Undetermined

Dataflow descriptions have been used in

a wide range of Digital Signal Processing (DSP)

applications, such as multi-media processing, and

wireless communications. Among various forms of

dataflow modeling, Synchronous Dataflow (SDF) is

geared towards static scheduling of computational

modules, which improves system performance and

predictability. However, many DSP applications

do not fully conform to the restrictions of SDF

modeling. More general dataflow models, such

as CAL (Eker and Janneck 2003), have been

developed to describe dynamically-structured DSP

applications. Such generalized models... (More)
Abstract in Undetermined

Dataflow descriptions have been used in

a wide range of Digital Signal Processing (DSP)

applications, such as multi-media processing, and

wireless communications. Among various forms of

dataflow modeling, Synchronous Dataflow (SDF) is

geared towards static scheduling of computational

modules, which improves system performance and

predictability. However, many DSP applications

do not fully conform to the restrictions of SDF

modeling. More general dataflow models, such

as CAL (Eker and Janneck 2003), have been

developed to describe dynamically-structured DSP

applications. Such generalized models can express

dynamically changing functionality, but lose the

powerful static scheduling capabilities provided by

SDF. This paper focuses on the detection of SDF-

like regions in dynamic dataflow descriptions—

in particular, in the generalized specification

framework of CAL. This is an important step for

applying static scheduling techniques within a dynamic

dataflow framework. Our techniques combine the advantages of different dataflow languages and tools,

including CAL (Eker and Janneck 2003), DIF (Hsu

et al. 2005) and CAL2C (Roquier et al. 2008). In

addition to detecting SDF-like regions, we apply

existing SDF scheduling techniques to exploit the

static properties of these regions within enclosing

dynamic dataflow models. Furthermore, we propose

an optimized approach for mapping SDF-like regions

onto parallel processing platforms such as multi-core

processors. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Journal of Signal Processing Systems
volume
63
issue
1
pages
129 - 142
publisher
Springer
external identifiers
  • scopus:79954601701
ISSN
1939-8115
DOI
10.1007/s11265-009-0445-1
language
English
LU publication?
yes
id
9cf79486-6d10-400c-b884-4426f82c0e66 (old id 2224876)
date added to LUP
2011-12-14 14:39:13
date last changed
2017-08-06 03:28:37
@article{9cf79486-6d10-400c-b884-4426f82c0e66,
  abstract     = {<b>Abstract in Undetermined</b><br/><br>
Dataflow descriptions have been used in<br/><br>
a wide range of Digital Signal Processing (DSP)<br/><br>
applications, such as multi-media processing, and<br/><br>
wireless communications. Among various forms of<br/><br>
dataflow modeling, Synchronous Dataflow (SDF) is<br/><br>
geared towards static scheduling of computational<br/><br>
modules, which improves system performance and<br/><br>
predictability. However, many DSP applications<br/><br>
do not fully conform to the restrictions of SDF<br/><br>
modeling. More general dataflow models, such<br/><br>
as CAL (Eker and Janneck 2003), have been<br/><br>
developed to describe dynamically-structured DSP<br/><br>
applications. Such generalized models can express<br/><br>
dynamically changing functionality, but lose the<br/><br>
powerful static scheduling capabilities provided by<br/><br>
SDF. This paper focuses on the detection of SDF-<br/><br>
like regions in dynamic dataflow descriptions—<br/><br>
in particular, in the generalized specification<br/><br>
framework of CAL. This is an important step for<br/><br>
applying static scheduling techniques within a dynamic<br/><br>
dataflow framework. Our techniques combine the advantages of different dataflow languages and tools,<br/><br>
including CAL (Eker and Janneck 2003), DIF (Hsu<br/><br>
et al. 2005) and CAL2C (Roquier et al. 2008). In<br/><br>
addition to detecting SDF-like regions, we apply<br/><br>
existing SDF scheduling techniques to exploit the<br/><br>
static properties of these regions within enclosing<br/><br>
dynamic dataflow models. Furthermore, we propose<br/><br>
an optimized approach for mapping SDF-like regions<br/><br>
onto parallel processing platforms such as multi-core<br/><br>
processors.},
  author       = {Gu, Ruirui and Janneck, Jörn and Raulet, Mickaël and Bhattacharyya, Shuvra S.},
  issn         = {1939-8115},
  language     = {eng},
  number       = {1},
  pages        = {129--142},
  publisher    = {Springer},
  series       = {Journal of Signal Processing Systems},
  title        = {Exploiting statically schedulable regions in dataflow programs},
  url          = {http://dx.doi.org/10.1007/s11265-009-0445-1},
  volume       = {63},
  year         = {2011},
}