Package jline.api.lossn

Class Lossn_links

java.lang.Object

jline.api.lossn.Lossn_links

public final class Lossn_links
extends Object

Link-level performance measures of a product-form loss network.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	Lossn_links.LossnLinksResult	Result of lossn_links(Matrix, Matrix, Matrix, String, Lossn_mci.Options).

Method Summary

Modifier and Type	Method	Description
static Lossn_links.LossnLinksResult	lossn_links(Matrix nu, Matrix A, Matrix C)	Per-link blocking and occupancy by the exact Manjunath-Sikdar transform.
static Lossn_links.LossnLinksResult	lossn_links(Matrix nu, Matrix A, Matrix C, String method)	Per-link blocking and occupancy with the given method.
static Lossn_links.LossnLinksResult	lossn_links(Matrix nuVec, Matrix Amat, Matrix cVec, String method, Lossn_mci.Options options)	Link-level performance measures of a product-form loss network.

Methods inherited from class java.lang.Object

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

Method Details

lossn_links

public static Lossn_links.LossnLinksResult lossn_links(Matrix nu,
 Matrix A,
 Matrix C)

Per-link blocking and occupancy by the exact Manjunath-Sikdar transform.

Parameters:

nu - offered load of route r (1xR)

A - capacity requirement of link j for route r (JxR)

C - capacity of link j (Jx1)

Returns:

E, Occ, Loss and QLen

lossn_links

public static Lossn_links.LossnLinksResult lossn_links(Matrix nu,
 Matrix A,
 Matrix C,
 String method)

Per-link blocking and occupancy with the given method.

Parameters:

nu - offered load of route r (1xR)

A - capacity requirement of link j for route r (JxR)

C - capacity of link j (Jx1)

method - 'exact' (default), 'rec', 'erlangfp', 'asy' or 'mci'

Returns:

E, Occ, Loss and QLen

lossn_links

public static Lossn_links.LossnLinksResult lossn_links(Matrix nuVec,
 Matrix Amat,
 Matrix cVec,
 String method,
 Lossn_mci.Options options)

Link-level performance measures of a product-form loss network.

For a loss network with offered loads nu, requirement matrix A and capacities C (A n <= C), returns for every link j and route r the probability E(j,r) that link j alone lacks the A(j,r) free units a route-r call needs, the mean occupancy of each link, and the route losses. With the exact methods the link blocking is

      E(j,r) = 1 - G(C - A(j,r) e_j) / G(C),
 

where G is the product-form normalization constant and e_j the j-th unit vector, computed from one extra constant per distinct (j, A(j,r)) pair. A route-r call is lost when ANY of its links blocks, so Loss_r is at least max_j E(j,r). The mean occupancy of link j is sum_r A(j,r) QLen_r.

Supported methods:

• 'exact' (or 'manjunath', 'ms'): Manjunath-style exact recursion
• 'rec': recursive decision diagram (Kirby-Wang style) normalization
• 'erlangfp': Erlang fixed point; E is the per-link Kaufman-Roberts blocking on the reduced loads
• 'asy': Kelly limit; E(j,r) = 1 - (1-a_j)^A(j,r)
• 'mci': Monte Carlo summation; E from the same samples as Loss

Parameters:

nuVec - offered load of route r (1xR)

Amat - capacity requirement of link j for route r (JxR)

cVec - capacity of link j (Jx1)

method - 'exact' (default when null or empty), 'rec', 'erlangfp', 'asy' or 'mci'

options - Lossn_mci options (samples, seed, alpha, gamma), used by 'mci'; may be null

Returns:

E, Occ, Loss and QLen