Abstract: To address the problem of phase wrapping in uniform circular array phase interferometers under Low-SNR and to enhance the robustness of DOA. A robust ambiguity resolution algorithm based on jump mode detection is proposed. The algorithm identifies two types of phase jumps induced by noise, systematically generates multiple candidate ambiguity sequences, and incorporates an adaptive pruning strategy for joint optimization. This approach transforms the phase unwrapping problem into a robust estimation of discrete jump patterns. Simulations results demonstrate that under the condition of a radius-to-wavelength ratio of 40 and a signal-to-noise ratio as low as −4 dB, the proposed algorithm maintains a direction finding success rate of 74%. Its performance is significantly superior to traditional fast unwrapping and correlation interferometer methods. The proposed algorithm effectively improves the success probability and robustness of phase unwrapping in low-SNR and wide-bandwidth environments. It provides a practical solution for highly robust direction finding in complex electromagnetic environments.