Path Loss Characterization Using Machine Learning Models for GS-to-UAV-Enabled Communication in Smart Farming Scenarios

Algorithm 2

ANN.

	Hyperparameter setting:
(1)	Let for RSSI training data sets and for the path loss training data set. The data set was separated as 80% for the training set and 20% for the test set.
(2)	Let two input layers for RSSI, where is the received power level and is horizontal distance between GS and UAV. For path loss, denotes two input layers, where is the path loss level.
(3)	One hidden layer and one output layer using RBF-based neural network (NN) were set, where the number of neurons .
(4)	The model of ANN is expressed by for RSSI and for path loss, where and represent the synaptic weights between the two input layers and a hidden layer for input RSSI training set and input path loss training set and denotes the connection weight between the neurons of the hidden and output layers.
	Model training:
(5)	The predicted value of RSSI is , and the model of path loss is , where and are the mean and standard deviation of a Gaussian function, respectively.