Class SolverMultiClosedTreeConvolution

java.lang.Object

solvers.SolverMulti

solvers.multiSolverAlgorithms.SolverMultiClosedTreeConvolution

public class SolverMultiClosedTreeConvolution
extends SolverMulti

Solves a multiclass closed model with the exact TreeConvolution algorithm, the tree based version of the convolution algorithm, which reduces the time and space requirements of the exact solution for models with large dimensions.

Author:

gcasale

Field Summary

Fields

Modifier and Type	Field	Description
protected int[]	population
protected QNModel	qnm	Contains the queueing network model

Fields inherited from class SolverMulti

classes, CLOSED_CLASS, clsNumJobs, clsRespTime, clsThroughput, DELAY, HOL, LD, LI, name, OPEN_CLASS, priorities, PRS, pw, queueLen, residenceTime, scQueueLen, scResidTime, scThroughput, scUtilization, servers, servTime, stations, sysNumJobs, sysResponseTime, sysThroughput, throughput, type, utilization, visits

Constructor Summary

Constructors

Constructor	Description
SolverMultiClosedTreeConvolution(int classes, int stations, int[] population)	Creates new SolverMultiClosedMonteCarlo

Method Summary

Modifier and Type	Method	Description
boolean	hasSufficientProcessingCapacity()	A system is said to have sufficient capacity to process a given load lambda if no service center is saturated as a result of the combined loads of all the classes.
boolean	input(int[] t, double[][][] s, double[][] v)	Initializes the solver with the system parameters.
void	solve()	Solves the system through the TreeMVA algorithm.

Methods inherited from class SolverMulti

getAggrQueueLen, getAggrQueueLen, getAggrResTime, getAggrResTime, getAggrThroughput, getAggrThroughput, getAggrUtilization, getAggrUtilization, getClsResTime, getClsResTime, getClsThroughput, getClsThroughput, getQueueLen, getQueueLen, getResTime, getResTime, getThroughput, getThroughput, getTotResTime, getTotThroughput, getUtilization, getUtilization, getVisits, input, input, printIndices, setServTime, toString

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Details

population

protected int[] population

qnm

protected QNModel qnm

Contains the queueing network model

Constructor Details

SolverMultiClosedTreeConvolution

public SolverMultiClosedTreeConvolution(int classes,
 int stations,
 int[] population)

Creates new SolverMultiClosedMonteCarlo

Parameters:

classes - number of classes

stations - number of service centers

population - array of class populations

Method Details

input

public boolean input(int[] t,
 double[][][] s,
 double[][] v)

Initializes the solver with the system parameters. It must be called before trying to solve the model.

Parameters:

t - array of the types (LD or LI) of service centers

s - matrix of service times of the service centers

v - array of visits to the service centers

Returns:

True if the operation is completed with success

solve

public void solve()

Solves the system through the TreeMVA algorithm. input(...) must have been called first.

Specified by:

solve in class SolverMulti

hasSufficientProcessingCapacity

public boolean hasSufficientProcessingCapacity()

Description copied from class: SolverMulti

A system is said to have sufficient capacity to process a given load lambda if no service center is saturated as a result of the combined loads of all the classes.
Must be implemented to create a multi class model solver.
WARNING: This method should be called before solving the system.

Specified by:

hasSufficientProcessingCapacity in class SolverMulti

Returns:

true if sufficient capacity exists for the given workload, false otherwise