Radio-to-Fiber Repeater Placement in Wireless Local-Loop Systems - Multiobjective Optimization Methodology: A Jumping Gene Approach - page 141

80

70

60

50

40

30

20

10

0

MICROGA MOGA PAES SPEA2 NSGA2 NSGA2 JG

Algorithm

Figure 6.5
Total. number. of. extreme. nondominated. solutions. for. each. MOEA. obtained. from. 10. sets. of.
testing.scenarios.

The.following.summarizes.the.results:

. 1.. As.compared.with.the.MOGA,.NPGA2,.and.MICROGA,.the.JG.was.

more.favorable.for.all.10.sets.

. 2.. As.compared.with.the.NSGA2,.the.JG.was.more.favorable.for.sets.

1,. 5,. 6,. 7,. and. 10.. However,. it. was. inconclusive. for. the. remaining.
sets.

. 3.. As.compared.with.the.SPEA2,.the.JG.was.more.favorable.for.sets.

4. and. 6-8.. Nevertheless,. it. was. inconclusive. for. the. remaining.
sets.

. 4.. As. compared. with. the. PAES,. the. JG. was. more. favorable. for. sets. 3.

and.10..However,.it.was.inconclusive.for.the.remaining.sets.

From.these.outcomes,.the.JG.scored.41.favorable.marks.and.19.inconclusive.
marks.for.all.10.testing.sets..As.a.consequence,.it.was.able.to.acquire.better.
sets.of.nondominated.solutions.than.other.MOEAs.with.good.convergence.
and.diversity.performance..Sample.nondominated.solution.sets.searched.by.
various.MOEAs.for.the.10.sets.of.testing.scenarios.are.depicted.in . Figure 6.6 .
for.reference.
In. conclusion,. based. on. different. performance. metrics,. JG. was. sta-
tistically. better. than. the. other. MOEAs. as. reflected. by. the. ε-indicator..
It. performed. better. in. convergence,. and. although. it. had. a. small. effect.
in. diversity,. it. obtained. more. extreme. cases. in. the. reference. Pareto.
optimal set.