Table 6 P-values from dunn’s post hoc analysis with bonferroni correction for comparing delay shift with other approaches.
From: Decentralized queue control with delay shifting in edge-IoT using reinforcement learning
Metric/Dataset | DRL | SEE-MTS | Delay-aware | Smart queue |
|---|---|---|---|---|
Mean RT/D4D | 0.004 | 0.009 | <0.001 | <0.001 |
Mean RT/Intel | 0.006 | 0.003 | <0.001 | <0.001 |
Mean RT/Edge | 0.005 | 0.012 | <0.001 | <0.001 |
Deadline/D4D | 0.001 | 0.006 | <0.001 | <0.001 |
Deadline/Intel | 0.004 | 0.008 | 0.001 | <0.001 |
Deadline/Edge | 0.007 | 0.010 | <0.001 | <0.001 |
Latency Std/D4D | 0.008 | 0.009 | <0.001 | <0.001 |
Latency Std/Intel | 0.005 | 0.004 | <0.001 | <0.001 |
Latency Std/Edge | 0.006 | 0.011 | <0.001 | <0.001 |
Recovery/D4D | 0.002 | 0.006 | <0.001 | <0.001 |
Recovery/Intel | 0.003 | 0.008 | <0.001 | <0.001 |
Recovery/Edge | 0.004 | 0.009 | <0.001 | <0.001 |
