Abstract: To address three major technical challenges in seismic exploration of ultra-deep fault-controlled carbonate reservoirs in area SHB of the Tarim Basin—strong absorption and attenuation by the thick desert overburden, energy masking by Permian igneous rocks, and interference from deep irregular Sangtamu intrusions—this study proposes a geometry design method based on Gaussian beam illumination analysis and dynamic forward-reverse illumination optimization. This methodology integrates complex igneous bodies and fracture-cavity reservoirs into a 3D geologic model. It then quantitativelyanalyze s target illumination under different acquisition parameter combinations using two indicators: mean illumination energy intensity ( \mu ) and uniformity index ( \theta ) , to rank parameter sensitivity and guide optimized survey design. A loop algorithm for dynamic forward-reverse illumination optimization is developed to identify blind zones via forward modeling and optimize the spatial layout of sources and receivers through reverse illumination. Synthetic and SHB field data tests demonstrate significantly improved illumination energy homogeneity in target layers, effectively addressing the challenge of insufficient imaging accuracy for ultra-deep fault-controlled reservoirs. This provides important technical support for the exploration and development of fracture-cavity carbonate reservoirs at depths exceeding 8,000 m in the Tarim Basin.