As a widely used over-the-top (OTT) video technique, adaptive bitrate video streaming has attracted considerable research attention and efforts. Traditional bitrate adaptation schemes rely on quality of service (QoS) to assess video delivery quality. However, in this paper, we argue that QoS may not accurately reflect the user-perceived video quality, especially in time-varying wireless networks. Instead, we adopt the concept of quality of experience (QoE) and propose a novel caching-based

more »

... ation scheme to improve the overall QoE, which can more comprehensively gauge the subjective user satisfaction of video quality. Experiments show that the proposed method outperforms existing adaptation schemes in terms of QoE in various network scenarios. This observation is mainly attributed to the merits of our scheme in properly leveraging channel bandwidth prediction and proxy-based content prefetching. Index Terms-Adaptive bitrate streaming, quality of experience (QoE), prefetching, bitrate adaptation, wireless networks