Virtual Machine Replica Placement Using a Multiobjective Genetic Algorithm

Table 4

Experimental results of heterogeneous environment

Model		Number of PMs	Distance	Performance degrading	F
	First-fit	2,066	22,559.00	556.00	32.61
	Next-fit	3,546	22,578.00	597.00	38.85
	Best-fit	3,403	18,750.00	604.00	37.35
gen 4,000; pop 25	NSGA-III	2,629.67  51.83	19,244.33  21.57	101.33  26.76	18.37
	PSO	2,839.33  3.21	20,319.00  54.53	550.00  19.67	34.23
	MOEA/D	2,610.00  27.22	18,934.67  14.98	183.00  9.64	20.90
gen 5,000; pop 50	NSGA-III	2,524.50  6.36	19,022  48.08	55  11.31	16.42
	PSO	2,824.00  4.24	20,156.00  70.71	560.00  1.41	34.45
	MOEA/D	2,566.50  9.19	18,906.50  12.02	181.00  12.73	20.68
gen 10,000; pop 50	NSGA-III	2,438.50  58.69	18,853.00  14.14	31.00  11.31	15.29
	MOEA/D	2,599.50  3.54	18,895.00  33.94	187.00  25.46	20.98

where bold numbers indicate the best result.