I have successfully used a spectral continuity constraint to dealias the least-squares slant stack operation. Two methods were used, the first applies a windowing constraint to the slant stack transform. This method works well for synthetic data but was too harsh for real data. The second method applies a weighting operator in the slant stack transform. The windowed slant stack can be used to predict and remove strong aliases before application of the weighted inverse slant stack. This method was effective in dealiasing a real dataset.
The method can be applied to a complete dataset or in local windows on the data. When choosing the window size there is a tradeoff between the effectiveness of the dealiasing and the ability to dealias curved events.