Federated learning (FL) is a machine learning framework that effectively helps multiple organizations perform data usage and machine learning models while meeting the requirements of user privacy protection, data security, and government regulations. However, in practical applications, existing federated learning mechanisms face many challenges, including system inefficiency due to data heterogeneity and how to achieve fairness to incentivize clients to participate in federated training. Due to this fact, we propose PFLIC, a novel personalized federated learning based on an iterative clustering algorithm, to estimate clusters to mitigate data heterogeneity and improve the efficiency of FL. It is combined with sparse sharing to facilitate knowledge sharing within the system for personalized federated learning. To ensure fairness, a client selection strategy is proposed to choose relatively “good” clients to achieve fairer federated learning without sacrificing system efficiency. Extensive experiments demonstrate the superior performance and effectiveness of the proposed PFLIC compared to the baseline.