To address the issues of a wide mainlobe width, insufficient interference suppression capability, and performance degradation in coherent interference environments associated with conventional beamforming algorithms, this paper proposes a null interference suppression technique based on beam sharpening, specifically including the H-LCMV and H-CTMV algorithms. This method divides a uniform linear array into left and right subarrays, computes the null-steering beam weight vectors for each subarray independently, enabling each subarray to form beam outputs with null characteristics. Based on the inherent phase delay between the subarrays, weight factors are constructed, and the outputs of the left and right subarrays are fused through nonlinear hyper-beamforming operations, thereby achieving significant mainlobe sharpening while deepening the null depth in the interference direction. This technique is applicable to both in coherent and coherent interference scenarios. Furthermore, by constructing a correction matrix to replace a single actual interference source with multiple virtual ones of equal intensity, the null is expanded into a notch with a certain width, effectively enhancing the algorithm′s robustness against slight perturbations in the interference direction. Simulation results and experimental data from the South China Sea show that, compared to traditional methods, the proposed algorithm significantly compresses the mainlobe width while achieving a null deepening of over 20 dB in the interference direction, effectively improving target resolution and interference suppression performance. Futhermore, the sharpened algorithm exhibits more precise pointing precision, effectively mitigating beam pointing deviation caused by ocean noise. These results fully validate the effectiveness of the proposed technique for target detection under strong interference and coherent source conditions, providing technical support for improving the detection performance of sonar systems in complex underwater acoustic environments.