Table 5 AICDQN improvement over baseline algorithms (Metric %).

From: Adaptive and intelligent customized deep Q-network for energy-efficient task offloading in mobile edge computing environments

Metric (%)

PGOA

ULOOF

DRL

DDPG

DDTO-DRL

Dropped tasks

79.68

79.03

48.00

79.83

62.65

Average delay

24.18

22.17

10.42

14.27

7.69

Energy consumption

18.60

15.32

13.22

5.41

0.94

Back to article page