Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling

Massimo Canala; Simone Casale Brunet; Endri Bezati; Marco Mattavelli; Jörn Janneck

doi:10.1109/SiPS.2014.6986054

Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling

Massimo Canala, Simone Casale Brunet, Endri Bezati, Marco Mattavelli, Jörn Janneck

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

Abstract

This paper presents a new approach to buffer dimensioning for dynamic dataflow implementations. A novel transformation applied to the execution trace graph of a dataflow program is introduced in order to generate an event driven system. It is shown how model predictive control theory techniques can be applied to such a system to analyse the execution space of a dataflow program and to define and to minimize a bounded buffer size configuration that corresponds to a deadlock free execution. Some experimental results obtained using two design examples, i.e. a JPEG and an MPEG HEVC decoder, are reported and compared to the state of the art results in order to show the effectiveness of the introduced approach.

Original language	English
Title of host publication	[Host publication title missing]
Publisher	IEEE - Institute of Electrical and Electronics Engineers Inc.
Pages	1-6
DOIs	https://doi.org/10.1109/SiPS.2014.6986054
Publication status	Published - 2014
Event	2014 IEEE Workshop on Signal Processing Systems - Belfast Duration: 2014 Oct 22 → 2014 Oct 24

Conference

Conference	2014 IEEE Workshop on Signal Processing Systems
Period	2014/10/22 → 2014/10/24

Subject classification (UKÄ)

Computer Science

Access to Document

10.1109/SiPS.2014.6986054

Cite this

@inproceedings{a58bac6cd70943cf8becc64fbdfbbb47,

title = "Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling",

abstract = "This paper presents a new approach to buffer dimensioning for dynamic dataflow implementations. A novel transformation applied to the execution trace graph of a dataflow program is introduced in order to generate an event driven system. It is shown how model predictive control theory techniques can be applied to such a system to analyse the execution space of a dataflow program and to define and to minimize a bounded buffer size configuration that corresponds to a deadlock free execution. Some experimental results obtained using two design examples, i.e. a JPEG and an MPEG HEVC decoder, are reported and compared to the state of the art results in order to show the effectiveness of the introduced approach.",

author = "Massimo Canala and {Casale Brunet}, Simone and Endri Bezati and Marco Mattavelli and J{\"o}rn Janneck",

year = "2014",

doi = "10.1109/SiPS.2014.6986054",

language = "English",

pages = "1--6",

booktitle = "[Host publication title missing]",

publisher = "IEEE - Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

note = "2014 IEEE Workshop on Signal Processing Systems ; Conference date: 22-10-2014 Through 24-10-2014",

}

Canala, M, Casale Brunet, S, Bezati, E, Mattavelli, M & Janneck, J 2014, Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling. in [Host publication title missing]. IEEE - Institute of Electrical and Electronics Engineers Inc., pp. 1-6, 2014 IEEE Workshop on Signal Processing Systems, 2014/10/22. https://doi.org/10.1109/SiPS.2014.6986054

Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling. / Canala, Massimo; Casale Brunet, Simone; Bezati, Endri et al.
[Host publication title missing]. IEEE - Institute of Electrical and Electronics Engineers Inc., 2014. p. 1-6.

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

TY - GEN

T1 - Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling

AU - Canala, Massimo

AU - Casale Brunet, Simone

AU - Bezati, Endri

AU - Mattavelli, Marco

AU - Janneck, Jörn

PY - 2014

Y1 - 2014

N2 - This paper presents a new approach to buffer dimensioning for dynamic dataflow implementations. A novel transformation applied to the execution trace graph of a dataflow program is introduced in order to generate an event driven system. It is shown how model predictive control theory techniques can be applied to such a system to analyse the execution space of a dataflow program and to define and to minimize a bounded buffer size configuration that corresponds to a deadlock free execution. Some experimental results obtained using two design examples, i.e. a JPEG and an MPEG HEVC decoder, are reported and compared to the state of the art results in order to show the effectiveness of the introduced approach.

AB - This paper presents a new approach to buffer dimensioning for dynamic dataflow implementations. A novel transformation applied to the execution trace graph of a dataflow program is introduced in order to generate an event driven system. It is shown how model predictive control theory techniques can be applied to such a system to analyse the execution space of a dataflow program and to define and to minimize a bounded buffer size configuration that corresponds to a deadlock free execution. Some experimental results obtained using two design examples, i.e. a JPEG and an MPEG HEVC decoder, are reported and compared to the state of the art results in order to show the effectiveness of the introduced approach.

U2 - 10.1109/SiPS.2014.6986054

DO - 10.1109/SiPS.2014.6986054

M3 - Paper in conference proceeding

SP - 1

EP - 6

BT - [Host publication title missing]

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 IEEE Workshop on Signal Processing Systems

Y2 - 22 October 2014 through 24 October 2014

ER -

Dataflow programs analysis and optimization using model predictive control techniques: An example of bounded buffer scheduling

Abstract

Conference

Subject classification (UKÄ)

Access to Document

Fingerprint

Cite this