An media aware platform for real-time stereoscopic video streaming adaptation

Stereoscopic video streaming is emerging as an important technology for future multimedia services. The delivery of real-time high quality three dimensional video to consumers remains a significant challenge due to the scarce resources of the access networks. While real-time rate adaptation has been proposed as a straight forward and easy to deploy method of minimizing transmission losses, its impact on the perceived three dimensional video quality is yet to be studied in depth. This paper investigates the impact of Media Aware Network Element (MANE) on the perceived quality of stereoscopic video measured by Peak Signal to Noise Ratio (PSNR) to both views. Instead of losing packets randomly during the video transmission, the proposed scheme selectively drops packets of the enhancement layers form both views, minimizing the distortion of both views. The Double Exponential Smoothing mechanism is selected for forecasting the transmission data rate. The experimental results indicate that by applying the proposed adaptation scheme, there is a significant improvement on the objectively measured quality of the received video.