A dataset for UWB ranging error mitigation in indoor environments, built using Decawave EVB1000 devices and the firmware contiki-uwb. Additional information can be found in the attached file "readme.txt" or in the paper Robust Ultra-wideband Range Error Mitigation with Deep Learning at the Edge. [ readme.txt ] * Every sample has the following structure: || CIR (157 float values) || || Error [m] || || Room (int) || || Obstacle (10 bool values) || || Measured Range (UWB) [m] || * Room encoding: 0 -> cross-room measurements 1 -> big room 2 -> medium room 3 -> small room 4 -> outdoor * Obstacle encoding: (1-hot encoding) 1000000000 -> wall 0100000000 -> polystyrene plate 0010000000 -> plastic (trash bin and chair) 0001000000 -> plywood plate 0000100000 -> cardboard box 0000010000 -> LCD TV 0000001000 -> metal plate 0000000100 -> wood door 0000000010 -> glass plate 0000000001 -> metal window * Reading Code: # Import libraries import pandas as pd import numpy as np # Extract dataset dataset = pd.read_pickle('dataset.pkl') # Select specific obstacle configurations ds = np.asarray(dataset.loc[dataset['Objects']=='011111111'][['CIR','Error']]) # Select specific rooms ds = np.asarray(dataset.loc[dataset['Room']==1][['CIR','Error']]) # Select all samples ds = np.asarray(dataset[['CIR','Error']]) # Get X,y for training X = np.vstack(ds[:,0]) y = np.array(ds[:,1])