QoE-driven Bandwidth Optimization

SSIMWave’s QoE-driven bandwidth optimization technology is at the core of SSIMWave’s Stream Optimizer (SSO) product.

Traditional video compression or encoding technologies attempt to achieve the best compromise between saving the bandwidth (bit rate) and minimizing the distortions. A common but major problem of such video encoders is the lack of a good definition of distortion. Indeed, most of them are optimized for mean squared error (or equivalently the peak signal-to-noise-ratio), which has been shown repeatedly to correlation extremely poorly with perceived video quality. As such, it is impossible for such video encoders to achieve the best performance when the visual quality of the video is concerned. What differentiates SSIMWave technology from all others is to let the most advanced video QoE assessment method (SSIMplus) to drive the bandwidth usage in a perceptually optimal manner.

SSIMWave’s bandwidth optimization technology operates as the brain of video encoders by guiding them to make the best use of the precious bandwidth while achieving the best possible video quality. An important fact worth mentioning is that the content type as well as the texture and motion complexity in a typical long video sequence change significantly over space and time, but the performance of video encoders varies largely with the type and complexity of the video content being encoded. As a result, using the same bandwidth to compress a long video sequence often lead to large quality variations over both space and time. Such a result is often disliked by end users due to unstable video quality. SSIMWave’s video QoE assessment technology provides the video encoders with the awareness of video quality variations deep to the pixel level. Subsequently, SSIMWave’s bandwidth optimization technology makes best use of such QoE awareness and redistribute the available bandwidth optimally.

The incorporation of SSIMWave’s bandwidth optimization technology in state-of-the-art video encoders results in many benefits. The most striking one is remarkable potential savings in the average bandwidth. The latest test on top of the best H.264 and H.265 open source video encoders show that adapting SSIMWave’s technology can lead to 20-60% bandwidth savings while maintaining the same level of video QoE. Another significant benefit is that the perceptual quality of the encoded video becomes constant overall time, regardless of the varying video content type and complexity. Furthermore, due to the powerful device-adaptation capability of SSIMWave’s QoE assessment technology, special video streams may be generated for optimal viewing experience for any given targeting viewing device.