High-speed computation of the Kleene star in max-plus algebra using a cell broadband engine

Authors:
Hiroyuki Goto;Takahiro Ichige
Affiliations:
Department of Management and Information Systems Engineering, Nagaoka University of Technology, Nagaoka, Niigata, Japan;Department of Management and Information Systems Engineering, Nagaoka University of Technology, Nagaoka, Niigata, Japan
Venue:
ACE'10 Proceedings of the 9th WSEAS international conference on Applications of computer engineering
Year:
2010

Citing 7
Cited 0

Introduction to Algorithms

Introduction to Algorithms
Simd Programming Manual for Linux and Windows (Springer Professional Computing)

Simd Programming Manual for Linux and Windows (Springer Professional Computing)
Introduction to the cell multiprocessor

IBM Journal of Research and Development - POWER5 and packaging
Brief paper: Input signal design for identification of max-plus-linear systems

Automatica (Journal of IFAC)
Programming the Cell Processor: For Games, Graphics, and Computation

Programming the Cell Processor: For Games, Graphics, and Computation
Efficient Calculation of the Transition Matrix in a Max-Plus Linear State-Space Representation

IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
An Efficient Solver for Scheduling Problems on a Class of Discrete Event Systems Using CELL/B.E. Processor

ISDA '09 Proceedings of the 2009 Ninth International Conference on Intelligent Systems Design and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

This research addresses a high-speed computation method for the Kleene Star of the weighted adjacency matrix in max-plus algebraic system. We focus on systems whose precedence constraints are represented by a directed acyclic graph (DAG), and implement on a Cell Broadband Engine™ (CBE) processor. Using the implementation on a Sony Playstation3™ (PS3) equipped with a CBE processor, we attempt to achieve a speedup by using two approaches: parallelization and SIMDization (Single Instruction, Multiple Data). We found that the SIMDization is effective regardless of the system's size and the number of processor cores used, and that the scalability of using multiple cores is remarkable especially for systems with large number of nodes.