The paper proposes an enhanced solution for improving the accuracy of underwater object positioning based on the Ultra-Short Baseline (USBL) sonar principle. In shallow water regions, complex hydrological phenomena affect the quality of underwater acoustic signal transmission and reception, making it essential to enhance the processing of these signal groups. Through the USBL system model, the paper presents a model capable of adaptively selecting accurate signals amidst the noisy underwater environment. The proposed model is implemented and validated in practice using FPGA hardware in Lan Ha Bay. The results demonstrate the effectiveness of the proposed algorithm compared to conventional approaches, as it extends the management range and optimizes the transmission power of the system, even in cases of low signal-to-noise ratio.