一种提高恶劣环境中电台话音质量的方法

针对恶劣环境中电台话音质量差这一难题，提出了一种有效的解决方法。首先在最小值控制的递归平均Ⅱ型（MCRA-2）算法基础上，提出了一种改进的MCRA-2算法。该算法采用自适应平滑参数和双检测门限，能有效减小音乐噪声，提高噪声估计准确性；然后将该噪声估计算法与对数最小均方误差估计器相结合，实现话音降噪，提高恶劣环境下电台话音质量；最后利用噪声估计结果和降噪后的话音信号，采用信噪比判决算法，自适应估计静噪门限，实现自适应静噪，进一步提高话音舒适度。实验表明，采用该方法的电台在大噪声、强干扰等恶劣环境中使用时，能明显改善话音质量，有效减轻收听者听觉疲劳。

A method to improve voice quality of radio in harsh environments

In view of the poor quality of radio voice in harsh environments,an effective solution is presented.On the basis of the minima-controlled recursive average II(MCRA-2) algorithm,an improved MCRA-2 algorithm is proposed.The algorithm uses adaptive smoothing parameters and double detection threshold to effectively reduce the musical noise and improve the accuracy of noise estimation.Then the noise estimation algorithm is combined with the logarithm minimum mean square error estimator to reduce the noise and improve the radio voice quality in harsh environments.Finally,by using noise estimation and voice signal after noise reduction,as well as the signal noise ratio decision algorithm,the quieting tone threshold is estimated adaptively,the adaptive quieting tone is realized,and the voice comfort is further improved.The experiment shows that the method can obviously improve the voice quality and reduce the hearing fatigue of the listener when the radio is applied in the environment of large noise and strong interference.