论文部分内容阅读
自适应预测器是32kb/s语声ADPCM系统中的一个关键部件,长期来人们一直致力于它的研究以解决精确性和复杂性的矛盾。本文从分析汉语语声的自相关函数着手,建立了语声自相关函数的数学模型,并根据这一模型提出了一种新型预测器——查表式自适应预测器。这种预测器根据不同的语声自相关函数直接从内存(EPROM)中读取预测系数,从而大大简化了自相关算法的运算步骤。本文中对以这种预测器构成的ADPCM系统作了性能分析,给出了计算机模拟结果。结果表明,用这种预测器构成的PCM-ADPCM系统的信噪比指标满足CCITT的建议G.712。
Adaptive predictor is a key component of 32kb / s speech ADPCM system. For a long time, people have devoted their research to solve the contradiction between accuracy and complexity. This paper starts with the analysis of the autocorrelation function of Chinese phonetics and establishes a mathematical model of phonetic autocorrelation function. Based on this model, a new predictor-look-up adaptive predictor is proposed. The predictor reads the prediction coefficients directly from the EPROM according to different speech autocorrelation functions, which greatly simplifies the operation steps of the autocorrelation algorithm. In this paper, the performance of ADPCM system constructed by this predictor is analyzed and the results of computer simulation are given. The results show that the signal to noise ratio of the PCM-ADPCM system constructed with this predictor meets the CCITT recommendation G.712.