The paper presents a correlation power analysis (CPA) method based on the Hamming distance model, which generalizes the Hamming weight model used in differential power analysis (DPA). The authors validate the model through experimental results and demonstrate its effectiveness in identifying the reference state and inferring secret keys in cryptographic devices. The CPA method is shown to be more robust and efficient compared to DPA, addressing issues such as "ghost peaks" and the limitations of DPA. The paper also discusses countermeasures against both CPA and DPA, highlighting that these methods share similar defense strategies. The conclusion emphasizes the advantages of CPA in terms of efficiency and robustness, while acknowledging the challenge of characterizing the leakage model parameters.The paper presents a correlation power analysis (CPA) method based on the Hamming distance model, which generalizes the Hamming weight model used in differential power analysis (DPA). The authors validate the model through experimental results and demonstrate its effectiveness in identifying the reference state and inferring secret keys in cryptographic devices. The CPA method is shown to be more robust and efficient compared to DPA, addressing issues such as "ghost peaks" and the limitations of DPA. The paper also discusses countermeasures against both CPA and DPA, highlighting that these methods share similar defense strategies. The conclusion emphasizes the advantages of CPA in terms of efficiency and robustness, while acknowledging the challenge of characterizing the leakage model parameters.