Content-Based Optimization for Subjective Quality of Dynamic Adaptive Streaming over HTTP

Focusing on the problem that traditional dynamic adaptive streaming over HTTP (DASH) only works on rate adaptation but performs nothing on optimizing consistent play-back quality and especially nothing on subjective quality, a content-based DASH adaptation decision is proposed.In the process of the decision taking, shots are regarded as adaptation units to be assigned safe factor and allocated network resource judged by its motion rank, and then a dynamic programming algorithm is employed to find the subjective quality maximized solution on how to download the next segments.The experimental results confirm that higher motion content achieve better quality because the radical need on network resource is satisfied, and that the lowest quality among all segments is increased, number of bit-rate switching results less, as well as no stall occurs.