Abstract: Based on the traditional detection methods have high detection error rate, energy consumption problems, puts forward the data radio network principal component analysis method and the average score combined with deviation calculation method of forgery attack detection method based on. Using function extreme value method to determine the degree of membership, to obtain the characteristics of radio network data forgery attack. Using the discrete wavelet transform method to pre process the data in the radio network. Using statistical analysis method, the determination of the radio network data forgery attack classification function, data forgery attack models for radio network, using principal component analysis method to the covariance matrix of the forgery attack on the initial radio network data eigenvalue decomposition, the average score calculation method is introduced to determine the deviation from the radio network data forgery attack deviation, realize the detection data forgery the radio network attack. Experimental results show that the improved detection method, the detection of energy consumption and false detection rate to a certain extent, compared with the traditional detection method is better, has a certain advantage.