A Deep Learning-Based User Selection Scheme for Cooperative NOMA System with Imperfect CSI

Algorithm 1

Data pretraining.

Input: Data .

Output: Data .

1: for is a sequence with all data of do

2: Separate the real part and the imaginary part of the plural data.

3: Set , where is the number of transmitted symbols in the sequence in theory.

4: The actual number of transmitted symbols is .

5: ifthen

6:   is trained by the proposed one-dimensional CNN.

7:  Extract feature vectors.

8: else

9:  Discard the data;

10: Update .

11: Output data .