Deep Reinforcement Learning-Based Joint Satellite Scheduling and Resource Allocation in Satellite-Terrestrial Integrated Networks

Table 2

Parameters of PSDDQN algorithm.

Parameters	Value
Batch size	32
Learning rate α	0.005
Number of leaves of SumTree structure	2000
Number of neurons in input layer	156
Number of neurons in hidden layers	64, 32
Number of neurons in output layer	24
Discount rate γ	0.9
Activation function	ReLU
Optimizer	Gradient descent optimizer