Package solvers.queueingnet.amva

package solvers.queueingnet.amva

Approximate MVA (AMVA) algorithms of JMVA for closed networks: Chow, Bard-Schweitzer, Linearizer, DeSouza-Muntz Linearizer and AQL (Aggregate Queue Length), as described in the JMVA chapter of the user manual.

Related Packages

Package	Description
solvers.queueingnet	Queueing network model of JMVA: stations, classes, service demands and visits used by the exact and approximate solution algorithms.
solvers.queueingnet.montecarloLogistic	Monte Carlo logistic regression used by the exact JMVA algorithms when the model dimension grows (logistic approximation of normalizing constants).
solvers.queueingnet.treeAlgorithms	Tree based exact algorithms of JMVA: TreeMVA and TreeConvolution, which reduce the complexity of the exact solution of large closed models.

Classes

Class	Description
AMVA	This class implements the abstract superclass for approximate MVA algorithms for solving multi-closed models
AQL	This class implements the AQL (Aggregate Queue Length) approximate MVA algorithm [ZES88] for solving multiclass closed models, as described in the AMVA section of the JMVA chapter of the JMT user manual.
BardSchweitzer	This class implements the Bard-Schweitzer approximate MVA algorithm [Sch79] for solving multiclass closed models.
Chow	This class implements the Chow approximate MVA algorithm [Cho83] for solving multiclass closed models.
Linearizer	This class implements the Linearizer [CN82] and De Souza-Muntz Linearizer [dSeSM90] approximate MVA algorithms for solving multiclass closed models, as described in the AMVA section of the JMVA chapter of the JMT user manual.
Solver	Solver is an abstract class, which contains the methods to initialize and solve models with a single class of jobs.
SolverMulti	SolverMulti is an abstract class, which contains the methods initialize and solve multiclass models.