Robust seismic-wave transmission of digital data in a deep geological environment

The use of seismic waves is well established in geophysics and geotechnical applications, particularly through surface-wave analysis, seismic reflection, and refraction methods. However, their use as a communication medium for data transmission remains uncommon, especially in deep geological environments where electromagnetic communication is strongly attenuated. However, this wireless communication method represents an interesting opportunity for underground infrastructures such as radioactive waste repositories, particularly for long-term monitoring. In this context seismic waves play an alternative role for long-term wireless underground communication. This study investigates the feasibility of transmitting digital information such as an image through geological layers using seismic waves. Experiments were conducted at the Andra Underground Research Laboratory (URL) to transmit encoded data for the first time from 490 m depth to the surface. The experimental setup consisted of 30 geophones deployed on the surface and two seismic source locations in the URL (borehole and on tunnel floor). The signals were generated from a depth of 490 meters to the surface by using a SeisMovie seismic source which is a low energy piezoelectric vibrator. The experimental setup generated stable and repeatable seismic signals, enabling reliable time-frequency analysis. Several encoding and decoding schemes were tested, including Hamming (7,4) code, frequency modulation, and Morse code. The analysis presented here focuses on the transmission of a 11x17 pixel image encoded using the Hamming (7,4) code. Each pixel was converted into a sequence of frequency activations and transmitted as seismic signals to the surface. The selected frequency band range from 103-124 Hz, corresponding to approximately 20-25 wavelengths for a P-wave velocity of 2500 m/s. Time-frequency analysis of the surface recordings enabled identification of the transmitted frequencies, which were then detected through threshold-based detection. Followed by Hamming decoding, this process successfully reconstructed the transmitted image. For a single transmission, only eight pixels out of 204 were incorrectly decoded, demonstrating the robustness of the encoding and detection workflow under realistic underground conditions. To assess repeatability and signal stability, the same experiment was repeated five times. Stacking the five recordings increased the signal-to-noise ratio by a factor of √5, significantly improving frequency detection. In the stacked case, only one pixel was incorrectly reconstructed, with 90% of the pixels decoded correctly without requiring error correction. The Hamming (7,4) code played an important role in correcting single-bit errors, particularly for individual transmissions with lower signal-to-noise ratios. The main limitation of the current workflow is the use of a fixed detection threshold, which does not fully account for amplitude variability due to noise or source performance. Future work will address these limitations by transmitting larger datasets and by improving detection robustness through an adaptive threshold. Overall, the results are promising but further work is needed before seismic waves can be considered a viable communication channel for transmission of information between deep geological environments and the surface.

Keywords: Seismic data transmission, subsurface monitoring, Hamming code, signal processing, noise reduction, signal detection, innovative technologies.