An Iterated Local Search Method for Cluster-Analysis-Based Multibaseline Phase Unwrapping

The accuracy of the cluster-analysis-based multibaseline phase unwrapping (CA-MBPU) algorithm is primarily influenced by the effective formulation and solution of intercept term constraint equations. However, the presence of local extrema during the discrete optimization process lead to suboptimal solutions, thereby compromising both accuracy and efficiency of the process. To overcome this issue, an improved solution strategy for the intercept term constraint equations is proposed. An iterated local search method integrating neighborhood operations is proposed in this paper, enabling dynamic adjustments to both step size and solution space. This enhancement improves the algorithm’s ability to explore the global solution space and prevents premature convergence to local extrema. Experiments results on both simulated and real data verified the effectiveness of the proposed approach.