Effective Capacity Maximization in beyond 5G Vehicular Networks: A Hybrid Deep Transfer Learning Method

Algorithm 2

Deep learning-based CUE-VUE matching.

1. Initialize statistical information of small-scale fading, locations of CUEs and VUEs, the realization of large-scale fading;

2. Padding virtual VUEs to the considered network;

3. Obtain the maximum traffic throughput sustained for each CUE-VUE pair from Algorithm 1;

4. Calculate the optimal matching scheme as training labels;

5. Deal with the training samples according to (32);

6. Deal with the training labels to reduce the sparsity;

7. Train the FNN parameters with data samples until the loss function converges;

8. Output the optimal model.