Appendices

AKey numbers

Values calculated in this report, with the chapter where each is derived. They are replaced by measured values as the work proceeds.

Table A.1.Principal calculated values.

BCode listings

Both listings are starting points for tasks T1.6 and T5.3 and have not yet been run on the hardware.

B.1Induction acquisition (48 kHz, codec as sound card)

import numpy as np, sounddevice as sd

FS, N = 48_000, 4096                           # 11.72 Hz bins, 85.3 ms blocks
BINS = np.array([16, 24, 37, 56, 85, 130, 197, 300, 456, 693, 1053, 1600])
PH = np.array([2.589, 5.287, 1.692, 2.610, 3.376, 2.940,
               1.805, 1.120, 0.966, 3.592, 5.042, 0.208])

n = np.arange(N)
block = np.cos(2 * np.pi * np.outer(n, BINS) / N + PH).sum(axis=1)
block *= 0.5 / np.abs(block).max()
tx = np.tile(block, 60)                                # about 5.1 s

# OUT 1 -> amplifier -> TX coil; IN 1 = TX current, IN 2-4 = RX coils, IN 5 = PPS
# the device name depends on the host driver
rec = sd.playrec(tx, FS, device="CS42448",
                 input_mapping=[1, 2, 3, 4, 5], output_mapping=[1])
sd.wait()

blocks = rec[N * 4:].reshape(-1, N, 5)                # drop start-up
spec = np.fft.rfft(blocks[:, :, :4], axis=1)[:, BINS, :]
H = (spec[:, :, 1:] / spec[:, :, :1]).mean(axis=0)    # 12 tones x 3 receivers
pps = np.flatnonzero(np.diff(rec[:, 4]) > 0.2)         # PPS edges, sample index

B.2Radar sweep with phase reference

import adi, numpy as np

sdr = adi.ad9361("ip:192.168.2.1")            # Pluto in two-channel mode
sdr.rx_enabled_channels = [0, 1]               # RX1 echo, RX2 coupled reference
sdr.sample_rate = int(2e6)
sdr.rx_buffer_size = 4096
sdr.tx_cyclic_buffer = True

fs, f_if, N = sdr.sample_rate, 100e3, sdr.rx_buffer_size
tone = np.exp(2j * np.pi * f_if * np.arange(N) / fs)
sdr.tx(0.5 * tone * 2**14)

def phasor(x):
    return np.vdot(tone, x) / N

S = []
for f in np.arange(500e6, 3000e6 + 1, 10e6):   # 251 steps
    sdr.tx_lo = sdr.rx_lo = int(f)
    for _ in range(2):
        sdr.rx()                                   # let the synthesisers settle
    echo, ref = sdr.rx()
    S.append(phasor(echo) / phasor(ref))

S = np.array(S)                                    # subtract a background sweep here
profile = np.abs(np.fft.ifft(S * np.hanning(len(S)), n=4096))

CGlossary

DReferences

1. [R1] Software defined radio for stepped-frequency, ground-penetrating radar. Thesis, Georgia Institute of Technology. repository.gatech.edu
2. [R2] Software defined radio for stepped-frequency, ground-penetrating radar. IEEE conference paper. IEEE Xplore 8128082
3. [R3] Software-defined ground penetrating radar using COTS hardware. GNU Radio Conference 2025. events.gnuradio.org
4. [R4] Ground-penetrating radar with HackRF software-defined radios. rtl-sdr.com
5. [R5] MetalMapper: a multi-sensor TEM system for UXO detection and classification. ResearchGate 235186078
6. [R6] Dipole models for UXO discrimination at live sites. SERDP/ESTCP project. serdp-estcp.mil
7. [R7] The spectral properties of the magnetic polarizability tensor for metallic object characterisation. arXiv:1906.00382
8. [R8] Characterisation of multiple conducting permeable objects in metal detection by polarizability tensors. arXiv:1809.08916
9. [R9] Detecting landmines for a safer world. Ingenia. ingenia.org.uk
10. [R10] MDS-10 dual sensor detector. Minelab. minelab.com
11. [R11] MDS-20 dual-sensor detector. EPE. epequip.com
12. [R12] Combining electromagnetic spectroscopy and ground-penetrating radar for the detection of anti-personnel landmines. Sensors 19(15):3390, 2019. mdpi.com
13. [R13] Deep learning-based classification of anti-personnel mines and sub-gram metal content in mineralized soil (DL-MMD). Scientific Reports, 2024. nature.com
14. [R14] Detection and classification of landmines using machine learning applied to metal detector data. ResearchGate 339640290
15. [R15] Landmine identification from pulse induction metal detector data using machine learning. ResearchGate 373286697
16. [R16] J. R. Wait, A conducting sphere in a time varying magnetic field. Geophysics 16(4), 666–672, 1951.
17. [R17] Cirrus Logic, CS42448 108 dB, 192 kHz 6-in, 8-out TDM CODEC, datasheet.
18. [R18] Analog Devices, ADALM-Pluto SDR active learning module and AD9363 documentation.
19. [R19] NATO Science for Peace and Security Programme: overview and Multi-Year Projects. nato.int; programme size and partner rules: Science|Business
20. [R20] Higher Education and Science Committee of Armenia, research grants. hesc.am
21. [R21] Foundation for Armenian Science and Technology, Advance Research Grants. fast.foundation
22. [R22] European Commission, list of participating countries in Horizon Europe. ec.europa.eu

EPhotographs

Photographs of the hardware are added as it is built. To add one, copy a file with the name shown into the images/ folder of the web server; it appears here without any other change.

FDocument history