Noise removal techniques for underground communication systems based on matching pursuit

In underground communication systems, continuous mud pulse signals are susceptible to pump noise during transmission, resulting in a high bit error rate (BER). In this paper, a Paradigm Inner Product Orthogonal Matching Pursuit (PIPOMP) algorithm is proposed for the transmission characteristics of continuous waves in the underground. First, the observation vectors of pump noise are obtained by signal cyclic prefix (CP) differencing, and the resulting observation vectors are more accurate than the traditional methods. Second, the columns of the sensing matrix that are most relevant to the observation vectors are selected as candidate support sets by computing the L2 paradigm. Then, the least squares method was used to solve for the estimated value of the pump noise at the previous moment. Finally, the pump noise is reconstructed by combining the correspondence between the time and frequency domains. This paper establishes a complete underground communication system. We simulate the denoising performance of pump noise under stable and unstable conditions and analyze the denoising performance of the PIPOMP algorithm in depth. Simulation results show that the algorithm significantly improves the interference immunity performance and reduces the system BER in the environment where pump noise interferes and the fading is more drastic.