JISE


  [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ] [ 6 ] [ 7 ] [ 8 ] [ 9 ] [ 10 ] [ 11 ] [ 12 ] [ 13 ] [ 14 ] [ 15 ]


Journal of Information Science and Engineering, Vol. 36 No. 6, pp. 1315-1337


Scheduling for the QOE Optimization of DASH Streaming


YU-CHI LIU1, MENG-HUANG LEE3,+, SHIN-HUNG CHANG4, WEI-CHUN CHUNG2,
KUAN-JEN WANG4, JAN-MING HO2 AND RAY-I CHANG1
1Department of Engineering Science and Ocean Engineering
National Taiwan University
Taipei, 116 Taiwan

2Institute of Information Science
Academia Sinica
Nankang, 115 Taiwan

3Department of Information Technology and Management
Shih Chien University
Taipei, 104 Taiwan

4Department of Computer Science and Information Engineering
Fu Jen Catholic University
New Taipei City, 242 Taiwan


DASH (Dynamic Adaptive Streaming over HTTP) is now the most widely used standard in video streaming. To support a DASH video transmission over a residential network with a small variation in bandwidth (such as a DSL-based network for IPTV), we designed an algorithm for generating the optimal transmission schedule, L2H. Given the transmission rate and initial delay, the schedule can optimize the quality of experience (QoE) metrics such as rebuffering, the lexicographically maximal resolution, the number of resolution switching events, and normalized average quality. We further present L2HB to consider the usage of the system buffer when applying L2H. L2HB demonstrates a benefit compared with algorithms proposed by previous studies based on objective and subjective QoE evaluations. In addition, by introducing a system buffer size constraint, the proposed algorithm can control the transmission schedule to enable the segments with the highest resolution to appear as soon as possible, encouraging viewers to continue watching.


Keywords: DASH, IPTV, quality of experience (QoE), scheduling, quality of service (QoS)

  Retrieve PDF document (JISE_202006_12.pdf)