Title :
A Polynomial Algorithm to Performance Analysis of Concurrent Systems Via Petri Nets and Ordinary Differential Equations
Author :
Zuohua Ding ; Yuan Zhou ; Mengchu Zhou
Author_Institution :
Lab. of Intell. Comput. & Software Eng., Zhejiang Sci-Tech Univ., Hangzhou, China
Abstract :
In this paper, a new method is proposed to evaluate the performance of concurrent systems. A concurrent system consisting of multiple processes that communicate via message passing mechanisms is modeled by a Petri net, which is in turn represented by a set of ordinary differential equations (ODEs) of a restricted type. The equations describe the system state changes, and the solutions, also called state measures, can be used for the performance analysis such as estimating response time, throughput and efficiency. This method can avoid a state explosion problem encountered by the conventional methods based on Continuous-Time Markov Chains. Its application to an IBM business system is given as an example.
Keywords :
Markov processes; Petri nets; concurrency control; differential equations; message passing; polynomials; IBM business system; ODE; Petri nets; concurrent systems; continuous-time Markov chains; efficiency estimation; message passing mechanism; ordinary differential equations; performance analysis; performance evaluation; polynomial algorithm; response time estimation; state measures; throughput estimation; Approximation methods; Equations; Mathematical model; Message passing; Performance analysis; Petri nets; Semantics; Continuous Petri net (CPN); continuous-time Markov chain (CTMC); discrete-event systems; ordinary differential equation; performance analysis; stochastic Petri net;
Journal_Title :
Automation Science and Engineering, IEEE Transactions on
DOI :
10.1109/TASE.2013.2285574