Feedback-Free Early VoIP Quality Adaptation Scheme in Next Generation Networks

The Next Generation Networks (NGNs) were proposed to provide ubiquitous access to multimedia services over the Internet. With the best effort nature of the Internet, the quality of real time multimedia services such as voice and video can rapidly degrade due to network impairments. In this paper, we propose a novel feedback-free Early VoIP Quality Adaptation Scheme (EVQAS) which can adapt voice send bit rate (e.g. Adaptive Multi-Rate (AMR) codec mode) automatically according to the buildup of queuing delay in the network. This differs from existing adaptation methods which adapt send bit rate according to either packet loss information or Perceived Quality of Service (PQoS) metrics (e.g. Mean Opinion Score (MOS)) derived from packet loss and delay. The proposed scheme can avoid congested packet loss which has an obvious impact on perceived voice quality. We implemented this adaptation mechanism in open Android mobile and tested in Open IMS Core testbed. Preliminary results show that the early adaptation mechanism can effectively release congestion before packet loss occurs. This has increased overall perceived voice quality when compared with existing adaptation method. This lightweight feedback-free early adaptation mechanism can be easily applied to other mobile devices to improve VoIP quality in current NGN VoIP applications.