中文摘要：

在多媒体通信中，一方面要求音频业务质量能适应不断变化的网络带宽，体现一定的层次性；另一方面，要求音频信源能够按码率任意截断从而更好地适应信道变化，以保证音频业务质量要求。这种信源的分层分级码流在网络传输时，更便于信道不对等保护技术的使用。提出了一种音频业务的可分层模型和基于EaacPlus核心编码器的可分层可分级音频编码方案。实验表明，分层模型的使用使得音质具有较强的层次感；可分级码率的增加能够带来音质的连续增强。

英文摘要：

Along with the variety of channel bandwidth, there are generally two requirements in the multimedia communication domain, especially for audio signal. One is that the quality of service （QoS） should be hierarchical, and the other is that the bit-rate of source coding should be arbitrarily adjusted for keeping a better QoS. It is evident that a scalable bit-stream with Mutli-Layer is convenient for Unequal Error Protection （UEP） of channel. This paper presents a layering model of audio transmission service and an audio coding scheme of adjustable bit- rate based on EaacPlus kernel. Some experiments show that proposed scheme can provide a real auditory feeling of hierarchical and continuously quality changing following the adjusting of bit-rate.