Abstract: To address the problems of the wide dynamic range of logging-while-drilling (LWD) resistivity data and the limited downhole storage and transmission bandwidth, a downhole data compression method based on logarithmic normalization was proposed, aiming to achieve the collaborative optimization of high precision and high compression efficiency. This method narrowed the value range of raw data through logarithmic transformation, eliminated the influence of physical units by combining normalization processing, and then mapped the data to a preset integer interval via dynamic range adjustment. It converted wide-range floating-point numbers into small-range integers using encoding technology, reducing the number of data bits while preserving the variation trend of data amplitude. Experimental results show that this method can reduce the number of compressed data bits by more than 50%, with the maximum relative error being only 2.85%, and the decompressed data is highly consistent with the raw data. By introducing a linear fitting function between the overall error coefficient and the scale mapping ratio, the influence law of the mapping ratio on compression effect is clarified, which enables flexible regulation of the balance between precision and storage according to actual demands. With low computational complexity, this method can be independently implemented for real-time processing by the downhole single-chip microcomputer, effectively improving the storage efficiency of downhole instruments and the surface data transmission rate. It provides a reliable technical solution for the efficient processing of LWD data with a wide dynamic range.