Frame dropping control by video content characteristics under the limited bandwidth

Due to the limited bandwidth and the unpredictable network, the MPEG-4 video coding stream with Fine Granularity Scalability (FGS) can be flexibly dropped by very fine granularity to adapt to the available network bandwidth. However, this is a trade-off problem between to drop the partial frames to reduce frame rate and to drop the partial packets of frames to keep frame rate. We propose the weighted assignment scheme in order to dynamically control frame dropping by our scheduling algorithm when the bandwidth is not sufficient. Our proposed scheme not only can integrate with our optimal scheduling algorithm to maximize total important packets by the available bandwidth but can also control frame dropping by video content characteristics. In our experimental results, we analyze the frame rate preference for video content characteristics under the limited bandwidth. Based on the prior information for each test sequence, our frame dropping control scheme can be better than other fixed frame rate.