Cross-tree adjustment for spatialized audio streaming over networked virtual environments

In recent years, integrated spatialized voice services have become an appealing application for networked virtual environments (NVE), e.g., Second Life. With a spatialized voice service, people can identify who is talking if there are several participants in the vicinity. The key challenge in a spatialized audio streaming application is how to disseminate audio streams while observing bandwidth limits of end-user computers and tight latency constraints, and it can be modeled as NP-complete problem. In this paper, we propose a heuristic algorithm called CTA for spatialized audio streaming over NVEs in a peer-to-peer manner. The proposed algorithm was applied to real avatar mobility traces collected from Second Life, and the simulation results demonstrate that (a) CTA can achieve a high ratio of successful to candidate receivers and (b) CTA enables most of the successful receivers to enjoy minimum latency.