barkincavdaroglu / Link-Prediction-Mesh-Network

PyTorch Implementation of a Deep Learning Model for Temporal Link Prediction in MANETs
2 stars 1 forks source link

Incorporate Map Features #34

Open barkincavdaroglu opened 1 year ago

barkincavdaroglu commented 1 year ago

Since SNR is affected by factors not present in graph topology, incorporating CNN for map image of area can be helpful. Basically we would 'convolve' parts of image per edge using GPS of node endpoints.

IMG_0041.jpg

The top part represents the graph (processed via GNN) and bottom part represents the image of the map (processed via CNN).

This is really hard to implement because we would have to factor in the path edge follows on map. Easier (but still hard) to accomplish with for loops, but infeasible as the dataset size increases. Harder to accomplish with vectors, but much faster.