Federated Learning: example dataset (FMCW 122GHz radars)

Open with (python code):

import scipy.io as sio database = sio.loadmat('data_base_all_sequences_random.mat')

The database contains 5 files:

• Data_test_2.mat: dimension 16000 x 512 - x_test = database['Data_test_2'] Contains 16000 FFT range measurements (512-point FFT of beat signal after DC removal) used for test database with corresponding labels in label_test_2.mat

• Data_train_2.mat: dimension 16000 x 512 - x_train = database['Data_train_2']
  Contains 16000 FFT range measurements (512-point FFT of beat signal after DC removal) used for training database with corresponding labels in lable_train_2.mat

• label_test_2.mat: dimension 16000 x 1 - y_test = database['label_test_2'] Contains the true labels for test data (Data_test_2.mat), namely classes (true labels) correspond to integers from 0 to 7: Class 0: human worker at safe distance >3.5m from the radar (safe distance) Class 1: human worker at distance (critical) <0.5m from the corresponding radar Class 2: human worker at distance (critical) 0.5m - 1m from the corresponding radar Class 3: human worker at distance (critical) 1m - 1.5m from the corresponding radar Class 4: human worker at distance (safe) 1.5m - 2m from the corresponding radar Class 5: human worker at distance (safe) 2m - 2.5m from the corresponding radar Class 6: human worker at distance (safe) 2.5m - 3m from the corresponding radar Class 7: human worker at distance (safe) 3m - 3.5m from the corresponding radar

• label_train_2.mat: dimension 16000 x 1 - y_train = database['label_train_2'] Contains the true labels for train data (Data_train_2.mat), namely classes (true labels) correspond to integers from 0 to 7: Class 0: human worker at safe distance >3.5m from the radar (safe distance) Class 1: human worker at distance (critical) <0.5m from the corresponding radar Class 2: human worker at distance (critical) 0.5m - 1m from the corresponding radar Class 3: human worker at distance (critical) 1m - 1.5m from the corresponding radar Class 4: human worker at distance (safe) 1.5m - 2m from the corresponding radar Class 5: human worker at distance (safe) 2m - 2.5m from the corresponding radar Class 6: human worker at distance (safe) 2.5m - 3m from the corresponding radar Class 7: human worker at distance (safe) 3m - 3.5m from the corresponding radar

• permut.mat (1 x 16000) contains the chosen random permutation for data partition among nodes/device and federated learnig simulation (see python code)