Indoor UWB CIR Data Set for Material Prediction

ABOUT

This data set contains spatially distributed CIR of multipath components in indoor environment acquired with ultra wideband (UWB) radio technology in microwave frequency range.
The data is labeled with the materials of the surfaces bounding the space (floor, ceiling, walls).
The data was collected for training and evaluating machine learning models for CIR-based indoor material prediction, but it may be also used for other studies based on indoor radio propagation data. 

AUTHORS

Teodora Kocevska, Andrej Hrovat, Tomaž Javornik

Department of Communication Systems

Jožef Stefan Institute, SI-1000 Ljubljana, Slovenia

DATA COLLECTION

The synthetic data is obtained using Remcom Wireless InSite v.3.3.3.
The CIR is estimated in 16,875 rooms in total. 
These rooms belong to 5,625 distinct room types and each room type is considered in three sizes. 
The number of distinct room types comes from the materials used for the floor, ceiling, and walls, having nine floor-ceiling material combinations and 625 wall-material combinations.

MATERIALS

•  floor: concrete, wood, floorboard
•  ceiling: concrete, plaster, wood
•  walls: brick, concrete, glass, plaster, wood

COMMUNICATION SYSTEM CONFIGURATION

Ultra wideband (UWB) radio technology is considered. The parameters of the communication system are set according to 802.15.4-2011** standard.
The configuration of system parameters is summarized as follows:

RADIO NODE POSITIONS

The data is collected using three acquisition layouts as follows:

1. Layout 1  

 Tx in the center of the room and Rx moved over uniform grid covering the room.

2. Layout 2

Tx in eight positions following circular pattern around the center of the room and Rx moved over uniform grid covering the room. 

The distance from the center of the room to the circumference of the circle is 0.5 m, and the spacing between the radio nodes is pi/4 rad.

3. Layout 3

Tx in four positions near the corners of the room (0.375 m from the walls) and Rx moved over uniform grid covering the room.

The corners of the grid are 0.25 m apart from the walls and the distance between the nodes is also 0.25 m.

Since the grid size is defined relatively to the room size, the total number of grid node positions is different in rooms with different sizes.

The total number of grid node positions is 121, 361, and 729 in S, M, and L rooms, respectively.

DATA ORGANIZATION

Data is saved in .csv files. Each file starts with a header line specifying the column names. 
The column names included in the .csv files are:

- Column 0: layout {center, circle, corners}
- Column 1: tx_point_id {1} for Layout 1, {1-8} for Layout 2, and {1-4} for Layout 3
- Column 2: rx_point_id {1-121} in S-rooms, {1-361} in M-rooms, and {1-729} in L-rooms
- Column 3: 1_phase_deg NUMERIC
- Column 4: 1_toa_ns NUMERIC
- Column 5: 1_power_dbm NUMERIC
- Column 6: 1_power_nw NUMERIC
- Column 7: 2_phase_deg NUMERIC
- Column 8: 2_toa_ns NUMERIC
- Column 9: 2_power_dbm NUMERIC
- Column 10: 2_power_nw NUMERIC
- Column 11: 3_phase_deg NUMERIC
- Column 12: 3_toa_ns NUMERIC
- Column 13: 3_power_dbm NUMERIC
- Column 14: 3_power_nw NUMERIC
- Column 15: 4_phase_deg NUMERIC
- Column 16: 4_toa_ns NUMERIC
- Column 17: 4_power_dbm NUMERIC
- Column 18: 4_power_nw NUMERIC
- Column 19: 5_phase_deg NUMERIC
- Column 20: 5_toa_ns NUMERIC
- Column 21: 5_power_dbm NUMERIC
- Column 22: 5_power_nw NUMERIC
- Column 23: 6_phase_deg NUMERIC
- Column 24: 6_toa_ns NUMERIC
- Column 25: 6_power_dbm NUMERIC
- Column 26: 6_power_nw NUMERIC
- Column 27: 7_phase_deg NUMERIC
- Column 28: 7_toa_ns NUMERIC
- Column 29: 7_power_dbm NUMERIC
- Column 30: 7_power_nw NUMERIC
- Column 31: 8_phase_deg NUMERIC
- Column 32: 8_toa_ns NUMERIC
- Column 33: 8_power_dbm NUMERIC
- Column 34: 8_power_nw NUMERIC
- Column 35: 9_phase_deg NUMERIC
- Column 36: 9_toa_ns NUMERIC
- Column 37: 9_power_dbm NUMERIC
- Column 38: 9_power_nw NUMERIC
- Column 39: 10_phase_deg NUMERIC
- Column 40: 10_toa_ns NUMERIC
- Column 41: 10_power_dbm NUMERIC
- Column 42: 10_power_nw NUMERIC
- Column 43: 11_phase_deg NUMERIC
- Column 44: 11_toa_ns NUMERIC
- Column 45: 11_power_dbm NUMERIC
- Column 46: 11_power_nw NUMERIC
- Column 47: 12_phase_deg NUMERIC
- Column 48: 12_toa_ns NUMERIC
- Column 49: 12_power_dbm NUMERIC
- Column 50: 12_power_nw NUMERIC
- Column 51: 13_phase_deg NUMERIC
- Column 52: 13_toa_ns NUMERIC
- Column 53: 13_power_dbm NUMERIC
- Column 54: 13_power_nw NUMERIC
- Column 55: 14_phase_deg NUMERIC
- Column 56: 14_toa_ns NUMERIC
- Column 57: 14_power_dbm NUMERIC
- Column 58: 14_power_nw NUMERIC
- Column 59: 15_phase_deg NUMERIC
- Column 60: 15_toa_ns NUMERIC
- Column 61: 15_power_dbm NUMERIC
- Column 62: 15_power_nw NUMERIC
- Column 63: room_size_surf_m2 {9} for S-rooms, {25} for M-rooms, and {49} for L-rooms
- Column 64: room_size_name {S} for S-rooms, {M} for M-rooms, and {L} for L-rooms
- Column 65: room_shape {square}
- Column 66: floor_mat {concrete, wood, floorboard}
- Column 67: ceiling_mat {concrete, plaster, wood}
- Column 68: wall1_mat {brick, concrete, glass, plaster, wood}
- Column 69: wall2_mat {brick, concrete, glass, plaster, wood}
- Column 70: wall3_mat {brick, concrete, glass, plaster, wood}
- Column 71: wall4_mat {brick, concrete, glass, plaster, wood}

Each row corresponds to separate radio link defined with the Tx and Rx nodes. 
It includes information about 
    (i)  the CIR-acquisition layout (position of the Txs and Rxs), 
    (ii)  the Tx and Rx, 
    (iii)  the CIR of 15 strongest multipath components, 
    (iv)  the room geometry, and 
    (v)  the materials of the surfaces bounding the space. 

- Column 0 specifies the layout. 
    The following maping is used: 
        Layout 1 -> center, 
        Layout 2 -> circle, and 
        Layout 3 -> corners.
- Column 1 specifies the Tx identifier.
- Column 2 specifies the Rx identifier.
- Columns 3-62 are the input attributes. 
    The input attributest represent the phase (in deg), ToA (in ns), received power (in dBm), and received power (in nW) for 15 strongest multipath components. 
    The column naming is X_Y_Z, where X is the multipath component identifier (1 to 15), Y is the propagation characteristic (phase, toa, or power), and Z is the unit (deg, ns, dbm, or nw).
- Column 63 specifies the surface of the room in square meters.
- Column 64 specifies the room size category (S, M, or L).
- Column 65 specifies the room-base shape.
- Columns 66-71 are the target attributes specifying the material of the floor, ceiling, wall 1, wall 2, wall 3, and wall 4, respectively. 

FOLDER STRUCTURE

The folder indoor_CIR_data contains:
    - one subfolder named CIR_data
        It contains three .csv files with CIR data named by the size of the rooms where the data is acquired.
        For naming the .csv files the following mapping is considered:
            - Small.csv  -> S-rooms
            - Medium.csv -> M-rooms
            - Large.csv  -> L-rooms
    - one subfolder named CIR_acquisition_details
        It contains .png file with schematic representation of the radio node positions considered for obtaining the data.
    - README.txt file

The folder structure is:
    - CIR_data
        - Small.csv
        - Medium.csv
        - Large.csv
    - CIR_acquisition_details
        - radio_node_positions.png
    - README.txt

REFERENCES

* R. sector of International Telecommunication Union (ITU-R), “Effects of building materials and structures on radio wave propagation above about 100 MHz,” International Telecommunication Union, ITU-R Recommendation P.2040-2, 2021.

** IEEE, “Standard for local and metropolitan area networks–Part 15.4: Low-rate wireless personal area networks (LR-WPANs),” IEEE, Standard IEEE 802.15.4-2011, 2011.