Doppler Spread Estimation Based on Machine Learning for an OFDM System

Table 1

Machine learning settings
Parameter	Value	Parameter	Value
(training sequence )	2000		1000
(input node ) in ML1	128000	(output node ) in ML1	100
(input node ) in ML2	500	(output node ) in ML2	100
Hidden and output layer	4	Epoch	2000
Neuron # in , , and	1000	Neuron in	100
Initial values of weights	Random	Dropout ratio	2%
Activation function	SoftMax (23)	Cost function	Cross entropy (24)