TY - GEN
T1 - Can accurate predictions improve video streaming in cellular networks?
AU - Zou, Xuan Kelvin
AU - Erman, Jeffrey
AU - Gopalakrishnan, Vijay
AU - Halepovic, Emir
AU - Jana, Rittwik
AU - Jin, Xin
AU - Rexford, Jennifer L.
AU - Sinha, Rakesh K.
PY - 2015/2/12
Y1 - 2015/2/12
N2 - Existing video streaming algorithms use various estimation approaches to infer the inherently variable bandwidth in cel- lular networks, which often leads to reduced quality of expe- rience (QoE). We ask the question: \If accurate bandwidth prediction were possible in a cellular network, how much can we improve video QoE?". Assuming we know the bandwidth for the entire video session, we show that existing stream- ing algorithms only achieve between 69%-86% of optimal quality. Since such knowledge may be impractical, we study algorithms that know the available bandwidth for a few sec- onds into the future. We observe that prediction alone is not sufficient and can in fact lead to degraded QoE. However, when combined with rate stabilization functions, prediction outperforms existing algorithms and reduces the gap with optimal to 4%. Our results lead us to believe that cellular operators and content providers can tremendously improve video QoE by predicting available bandwidth and sharing it through APIs.
AB - Existing video streaming algorithms use various estimation approaches to infer the inherently variable bandwidth in cel- lular networks, which often leads to reduced quality of expe- rience (QoE). We ask the question: \If accurate bandwidth prediction were possible in a cellular network, how much can we improve video QoE?". Assuming we know the bandwidth for the entire video session, we show that existing stream- ing algorithms only achieve between 69%-86% of optimal quality. Since such knowledge may be impractical, we study algorithms that know the available bandwidth for a few sec- onds into the future. We observe that prediction alone is not sufficient and can in fact lead to degraded QoE. However, when combined with rate stabilization functions, prediction outperforms existing algorithms and reduces the gap with optimal to 4%. Our results lead us to believe that cellular operators and content providers can tremendously improve video QoE by predicting available bandwidth and sharing it through APIs.
UR - http://www.scopus.com/inward/record.url?scp=84942425262&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84942425262&partnerID=8YFLogxK
U2 - 10.1145/2699343.2699359
DO - 10.1145/2699343.2699359
M3 - Conference contribution
AN - SCOPUS:84942425262
T3 - HotMobile 2015 - 16th International Workshop on Mobile Computing Systems and Applications
SP - 57
EP - 62
BT - HotMobile 2015 - 16th International Workshop on Mobile Computing Systems and Applications
PB - Association for Computing Machinery, Inc
T2 - 16th International Workshop on Mobile Computing Systems, HotMobile 2015
Y2 - 12 February 2015 through 13 February 2015
ER -