Table 8.

Time consumed by stochastic channel-based federated learning with pruning using different pruning rates for each loop or different total pruned proportions.

Pruning parameter		Time (seconds)
Pruning rate/loop
	10%	11,144
	20%	8561
	30%	11,852
	40%	8389
	50%	12,000
Total pruned
	10%	25,755
	19%	22,717
	27%	17,579
	34%	15,909
	41%	8050
	47%	8469