本文提出了一种能够在全球移动通信系统(Global System for Mobile Communications,GSM)中使用的信息隐藏算法——基于分析合成的能量比调整(Analysis-By-Synthesis Energy Ratio Adjust,AERA)算法.算法采用了分析合成(Analysis-By-Synthesis,ABS)技术,在嵌入过程中根据输入明文语音实时的调整嵌入强度,使得隐藏效果和解码效果都达到最佳值.大量仿真试验结果表明算法对GSM中的规则脉冲激励长期预测编码(Regular Pulse Excited-Long Term Prediction Coding,RPE-LTP Coding)、自适应差分脉冲编码调制(Adaptive Differential Pulse Code Modulation,ADPCM)等语音压缩编码以及滤波操作有很强的鲁棒性,携密语音的分段平均信噪比达到37dB,可以达到透明性要求.算法简单易行并且是基于盲检测的,具有很高的实用性.

This paper presents a speech-hiding algorithm-Analysis-By-Synthesis Energy Ratio Adjust Algorithm (AERAA) that can be deployed on Global System for Mobile Communications (GSM).The algorithm is based on the technique of Analysis-By-Synthesis (ABS) and can adjust adaptively the strength of the security information component according to public speech.The hiding capacity of the algorithm is 50 bits per second.After a lot of experiments,it is proved that this algorithm has strong robustness to many attacks such as Regular Pulse Excited-Long Term Prediction coding (RPE-LTP coding) on GSM,Adaptive Differential Pulse Code Modulation (ADPCM) compression and most of filters.The average Signal Noise Ratio (SNR) of the middle speech is above 37db,therefore the imperceptibility can be ensured.Moreover the algorithm is based on blind-detection,it is simple and effective.