Base Station Placement in WLANs - Multiobjective Optimization Methodology: A Jumping Gene Approach - page 188

110
100
90
80
70
60
50
40
30
20
10
0

MICROGA PAES MOGA NPGA2 SPEA2 NSGA2 JG

Algorithm

Figure 8.5
Total. number. of. extreme. nondominated. solutions. for. scenario. b:. 80. dB.. (From. Tang,. K.. S.,.
Kwong,.S.,.Man,.K..F.,.A.jumping.genes.paradigm:.Theory,.veriication.and.applications, IEEE
CircuitsandSystemsMagazine ,.8(4),.18-36,.2008.)

8.7.4 Statistical Test using the binary ε -indicator

Table 8.6 . shows. the. statistical. results. for. the. binary. ε-indicator. in. terms.
of. the. number. of. occurrences. of. three. comparison. cases. (same. as. those.
given.in.Section 7.8.4).for.the.two.scenarios.
In.conclusion,.the.JG.was.more.favorable.than.other.MOEAs.for.both.sce-
narios..It.obtained.better.sets.of.nondominated.solutions.with.better.conver-
gence. and. diversity. performance.. Sample. sets. of. nondominated. solutions.
searched.by.different.MOEAs.for.scenarios.a.and.b.are.shown.in . Figure 8.6 .
to . Figure 8.9 . for.reference.

Table 8.6
Statistical.Results.of.Binary.ε-Indicator.in.Terms.of.the.Number.of.Occurrences.
in.Three.Different.Cases.for.Scenarios.a.(90.dB).and.b.(80dB)

Scenarios Case MOGA NPGA2 NSGA2 SPEA2 PAES MICROGA

a:.90.dB Case.I 1,706 1,912 1,018 1,402 2,138 2,482

Case.II 137 82 642 366 16 0

Case.III 657 506 840 732 346 18

b:.80.dB Case.I 1,435 1,513 956 1,217 1,769 2,477

Case.II 143 131 860 297 0 14

Case.III 922 856 684 986 731 9

Source: Data.from.Chan,.T..M.,.Man,.K..F.,.Tang,.K..S.,.Kwong,.S.,.A.jumping-genes.para-

digm. for. optimizing. factory. WLAN. network, IEEE Transactions on Industrial
Informatics ,.3(1),.33-43,.2007.