Skip to main content
. 2025 Sep 26;15:33122. doi: 10.1038/s41598-025-18353-8

Table 2.

Performance metrics for RMAL-Based Multi-UAV system.

Metric Value/Observation Description
Disk Radius Inline graphic 600 m The radius of the area where ground users are randomly distributed.
Number of Users Inline graphic 80 Total number of ground users uniformly dispersed across the disk
UAV Inline graphic 80 m Altitude at which UAVs are operating.
Noise Power Inline graphic −70 dBm Assumed noise power in the simulation.
Sub-Channel Bandwidth (Inline graphic) 65 kHz Bandwidth allocated to each sub-channel.
Time Slot Duration (Inline graphic) 0.1 s Duration of each time slot in the simulation.
Carrier Frequency (Inline graphic) 2 GHz Frequency used in the simulation.
Maximum UAV Power (Inline graphic) 23 dBm Maximum power output per UAV in the simulation.
Power Levels (Inline graphic) 3 Number of discrete power levels for UAVs.
Minimum SINR (Inline graphic) 3 dB Required minimum SINR for maintaining communication quality.
Average Reward (Inline graphic) Increase unit t = 2 250 time slots Reflects the ability of the RMAL algorithm to maximize long-term rewards before UAVs leave the disk.
Immediate Reward (Inline graphic) Increase with iterations; plateaus after Inline graphic time slot Represents the instantaneous rewards received per time slot.
Efficiency Plateau Observed at Inline graphic seconds Rewards stop increasing as UAVs exit the operational area.
Execution Time Efficiency RMAL shows consistent reward improvement over < 250 time slots Efficient in achieving rewards quickly due to adaptive Q-learning updates.