Class ConvolutionCore

java.lang.Object
solvers.queueingnet.treeAlgorithms.treeConvolution.ConvolutionCore
public class ConvolutionCore extends Object
Implements the lowest level convolution operations, as outlined in the LamLien83 paper.
Author:
Ben Homer, 2014.

  • Field Details

    • qnm

      protected QNModel qnm
      The queueing network we are trying to solve.

  • Constructor Details

    • ConvolutionCore

      public ConvolutionCore(QNModel qnm, ClassCoverageUtils ccu)
      Constructor.
      Parameters:
      qnm - The queueing network we are trying to solve.
      ccu - A ClassCoverageUtils instance.

  • Method Details

    • computeLeafNodeGArray

      public void computeLeafNodeGArray(Node n, PopulationVector popVec, boolean backup)
      Computes the g-array at the specified leaf node by implementing Eq. (9) from Lam/Lien paper.
      Parameters:
      n - The leaf node.
      popVec - The population vector to use during the computation.
      backup - Whether to store the results in a backup g-array map.

    • leafNodeDelayInitalCondition

      protected BigDecimal leafNodeDelayInitalCondition(PopulationVector p, PopulationVector popVec, Node n)
      Computes the g-array at the specified leaf node representing a delay station.
      Parameters:
      p - the p value
      popVec - The population vector to use during the computation.
      n - The leaf node representing a delay station.
      Returns:
      the resulting big decimal

    • computeSubnetNodeGArray

      public void computeSubnetNodeGArray(Node node, PopulationVector popVec, boolean backup)
      Computes the g-array at the specified subnet (non-leaf) node.
      Parameters:
      node - The subnet node.
      popVec - The population vector to use during the computation.
      backup - Whether to store the results in a backup g-array map.

    • feedbackFilter

      public void feedbackFilter(Node parentNode, PopulationVector parentPcMax, PopulationVector popVec, boolean backup)
      Implements equation (6) from Lam/Lien for combining two leaf node stations, more efficient than normal convolution if one is a fixed-rate station.
      Parameters:
      parentNode - The node to compute the g-array for.
      parentPcMax - The parent population max for partially covered classes.
      popVec - The max population vector to use in the calculation.
      backup - Whether to store the results in a backup g-array map.