The rapid development of mobile communication technology has generated abundant unlabeled radio source signals. To fully utilize unlabeled data, this paper proposes an Independence Criterion-based Unsupervised Domain Adaptation(ICUDA) method for specific emitter identification. The independence criterion is employed to measure the similarity between the source domain and the target domain, and combined with an improved convolutional neural network to transfer knowledge from the source domain to the target domain, thereby helping improve the classification performance of the target domain that contains only unlabeled data. Under seven transfer scenarios constructed based on Software-Defined Radio(SDR) dataset collected in a laboratory environment, compared with baseline methods and three unsupervised domain adaptation methods, the proposed method achieves the best classification performance in the target domain across all scenarios, with an average recognition accuracy of 84.2%, demonstrating that the proposed method can extract features with good inter-class separability and intra-class compactness on the target domain, effectively reducing the target domain's dependence on high-quality labeled data.