To solve the problem that radar emitter signal feature evaluation is not objective and lacks an evaluation basis, a group evaluation method based on interval fuzzy principle, cross-entropy, and multi-criterion compromise method is proposed. First, establish a signal noise rate (SNR) evaluation model, then combine the triangular fuzzy number cut set and interval approximation ideas to convert the evaluation system of radar emitter signal features into interval-valued intuitionistic fuzzy numbers. Next, optimize the SNR weight based on Hamming distance, and use interval-valued intuitionistic fuzzy weighted average operator-integrated group decision matrix based on the maximum entropy method to calculate the attribute weights. Finally, the multi-criteria compromise method and fuzzy cross-entropy are used to achieve the feature scheme ranking. Simulation results show that the proposed method can provide a consistent feature evaluation-ranking scheme with the actual situation and is consistent with the analysis results of the approach to the ideal point method and projection model method, verifying the feasibility and effectiveness of the proposed method.