Distribution and Rendering of Real-Time Scene Based on the Virtual Reality System - Advances in Image and Graphics Technologies - page 201

The testing machine starts the video streaming at the rate of 300 Kb/s when real-time
video is uploaded. The video is recorded 60 second. The delay time is counted by
HTC Onex's request for the live streams when the pressure request number is 100,
150, 200, 250 or 300.
Analysis: As shown in figure 6, memory factor is below 0.5.Bandwidth factor is
the key problem for the data request. When the mobile client is uploading video
frames, the procedure of frame processing and saving will lead to a delay of 1s so that
the delay is more than 2s at least. When the number of pressure test connections of the
delay figure of subnet one is under 200, the delay time is about 2 second. When the
pressure test connection number is more than 200, the bandwidth of the server has
been filled so that the delay time is increased because there is only a distributed server
in subnet one. However, there are enough bandwidths for second subnet so that the
delay time is still 2s.

6 Conclusions

In this research we have presented a system that includes video recording, transmis-
sion, distribution and rendering based on the design and implementation of large-scale
distributed virtual reality system. The real-time video scenes through mobile client are
browsed or projected at the alignment angle position in the virtual reality. The data
cache replacement policy and load balancing policy is designed for the content distri-
bution server. In the real application environment, delay is increased compared to the
video chat application because of the architecture of http server transmission as well
as the delay of transcoding. Meanwhile, the live video scene is only displayed and
projected in the virtual scene at the alignment position in the current system. It's bet-
ter to use a good algorithm to blend virtual scene with real scenes. We believe that all
these questions are worth a continued exploration in the future studies.

Reference

1. Wang, G.: ViWoSG: Ultra-large-scale Distributed Virtual Environment. SCIENCE
CHINA Information Sciences 39(1), 96-106 (2009)
2. Caudell, T.: AR at boeing (EB/ OL) (1990),
http://www.ipo.tue.nl/homepages/mrauterb/presentations/HCI2h
istory/tsld096.html
3. Quinn, B., Ajmeroth, K.: IP multicast applications: Challenges and Solutions (EB/OL),
http://www.ietf.org/rfc/rfc3170.txt (February 20, 2011)
4. Tran, D.A., Hua, K.A., et al.: A Peer-to-peer Architecture for Media Streaming. IEEE
Journal on Selected Areas in Communications 22(1) (2004)
5. Vakali, A., et aI.: Content Distribution Networks Status and Trends. IEEE Internet Compu-
ting, 68-74 (November 2003)
6. Mulerikkal, et aI.: An Architecture for Distributed Content Delivery Network. IEEE
(2007)