一种面向多点交互的自适应音频混合方法

An Adaptive Audio Mixing Method Oriented Multi-point Interaction

现有的全分布式混合通信模型和集中混合通信模型在网络带宽动态变化时很难满足多点交互的声音质量和可扩展性。鉴于此,论文提出了一种分层通信模型和基于此模型的自适应音频混合方法。这种音频混合方法根据当前网络可用带宽以及应用程序的QOS要求,自适应调整音频混合的程度。该方法已经在同步实时教学系统中得到了应用。相关数据表明,在IP网上,自适应音频混合方法可以有效地提高多点交互的声音质量和可扩展性。

The current audio communication models,such as peer to peer or centralized mixing,are hard to satisfy the quality of audio and extensibility of multi-point interaction when the network bandwidth changes dynamically.To solve these issues,a hierarchical communication model and an adaptive audio mixing method based on the model are proposed in the paper.The method can adjust the mixing degree of audio streams automatically according to the current available bandwidth and the QoS requirements.It has been applied in the synchronous real-time learning system.Collected datas show that it can effectively improve the quality and extensibility of multi-point audio interaction in IP networks.