In basic framework of ICA,the assumption of independence is crucial for the identifiability of the model,however,in practiceal application the source signals often have some relationship.

• 独立分量分析(ICA)模型中,有关独立性的假设对模型的可识别性至关重要,然而在实际应用中信号源之间往往有某种程度的联系。