HTTP-based Adaptive Streaming (HAS) plays a key role in over-the-top video streaming. It contributes towards reducing the rebuffering duration of video playout by adapting the video quality to the current network conditions. However, it incurs variations of video quality in a streaming session because of the throughput fluctuation, which impacts the user's Quality of Experience (QoE). Besides, many adaptive bitrate (ABR) algorithms choose the lowest-quality segments at the beginning of the streaming session to ramp up the playout buffer as soon as possible. Although this strategy decreases the startup time, the users can be annoyed as they have to watch a low-quality video initially. In this paper, we propose an efficient retransmission technique, namely H2BR, to replace low-quality segments being stored in the playout buffer with higher-quality versions by using features of HTTP/2 including (i) stream priority, (ii) server push, and (iii) stream termination. The experimental results show that H2BR helps users avoid watching low video quality during video playback and improves the user's QoE. H2BR can decrease by up to more than 70% the time when the users suffer the lowest-quality video as well as benefits the QoE by up to 13%

Cloud Gaming (CG) solutions run games on cloud servers and stream the compressed video of the game over a network. While this has many benefits for both users and game companies, CG imposes two major challenges: low delay and high bandwidth are essential to ensure QoE. While the tolerable delay varies based on state of the game, previous researches do not fully exploit this concept to enhance QoE. This paper demonstrates the importance of the game-state as a criterion for game requirements. Accordingly, a game-State Aware Resource Allocation for Software-Defined Network (SDN)-based cloud gaming (SARA-SDN) is presented, which incorporates the game state into decision making. First, an SDN architecture is designed to facilitate communication of game information and control commands between game-server, SDN controller, and network switches. Second, a resource allocation algorithm is proposed that dynamically considers game-specific requirements such as delay, bandwidth, and game-state. Finally, SARA-SDN is implemented and evaluated in a simulated environment, where real users played in real time, and different aspects of QoE was measured via extensive subjective tests. Results show the superiority of SARA-SDN over existing methods (on average 1.26, 1.18, 1.21, and 1.12 increase in QoE, responsiveness, controllability, and immediate feedback, respectively, in a scale of 0-5).

Omnidirectional MediA Format (OMAF), a recent standard for representation of 360° video content, supports only equirectangular projection (ERP) and cubemap projection (CMP) and their region-wise packing with a 4K decoding-capability constraint on the end-user devices. Streaming of 4K content allows only a limited viewport resolution, which is lower than the resolution of many current head-mounted displays (HMDs). Two 6K-effective multi-resolution packing schemes for ERP and CMP are recommended by the OMAF standard, which take full advantages of HMDs with quad HD resolution while sticking with 4K-decoding constraint. This paper studies and compares the performance of these two packings in the context of viewport-adaptive streaming using the High Efficiency Video Coding (HEVC) standard and Zonal-PSNR test methodology. Overall, the experimental results show that rate-distortion performance of both packings sequence wise mixed but similar in average.

In recent years, cloud gaming (CG)'s market share has grown in both academic research and adoption by the industry. Despite its advantages, CG has its own challenges. Delay has one of the most detrimental effects on the QoE of players. This is more intense in multiplayer cloud gaming (MCG), where players interact, and the game conditions should be the same for all players to be able to play fairly. Hence, network resources should be allocated to players based on their requirements and current network conditions, so that inter-delays of interacting players would be relatively close. To remedy this and to improve the QoE of CG, this paper combines the two concepts of edge computing and software-defined networks (SDN). Edge Computing brings servers closer to clients to decrease end-user delay. SDN enables dynamic management of network, based on user requirements and network states. We propose a suboptimal dynamic server and path selection method that allows players to achieve their required delay and bandwidth. Moreover, this method minimizes the inter-player delays for players in the same group. Experimental results show that our resource provisioning reduces inter-player delay and total delay by 94%, 31%, respectively, compared to competing methods.

DASH streaming for 360-degree content is gaining more traction in research and commercial areas. The usefulness of H.265/HEVC Motion Constrained Tile Sets (MCTS) and viewport-dependent streaming is recognized. Streaming high quality tiles for the viewport and lower quality tiles in the surrounding area provide considerable bandwidth saving. However, required bit rate estimation is difficult for a changing viewport with variable tiles, prompting research in development of new ABR algorithms. In this paper, we present Full Sphere Bit rate Estimation (FriSBE), an Adaptive Bit Rate (ABR) technique for tiled omnidirectional content that reuses existing ABR algorithms developed for 2D video. We implemented FriSBE in an OMAF-based DASH player and found that it is able to adapt to both head motion and changing network conditions with minimal stalls. Our experiments showed no stall events for stable conditions with or without head motion, and on average 1s stall duration for varying network conditions when bandwidth was sufficient for at least the lowest quality video, despite a short 3 second buffer.

Subblock-based affine motion model and prediction refinement with optical flow for affine (PROF) enhances support for more complex motions in Versatile Video Coding (VVC). Current syntax for affine motion model supports signaling large rotation (e.g. close to 90/180/270 degree) and flipping, however, engine of affine motion compensation and PROF does not work optimally for these cases. In case of larger rotation, the predicted subblock is not useful and also there is a big discontinuity in the borders of predicted subblock. Furthermore, PROF increases discontinuity. In this study we propose a solution to detect large rotation, harmonize PROF process, and apply post processing to subblock prediction samples. In video-based point cloud coding (V-PCC) use case, where YUV sequences are generated based on point cloud data, -2.81% and -3.12% overall Bjøntegaard Delta bitrate gain is achieved for random-access and low-delay configurations, respectively. However, the method does not have noticeable effect on encoding and decoding time.