Dynamic Adaptive Streaming over HTTP (DASH) is a promising scheme for improving the Quality of Experience (QoE) of users in video streaming. However, the existing schemes do not perform coordination among clients and depend on fixed heuristics. In this paper, we propose an adaptive streaming scheme with reinforcement learning in edge computing environments. The proposed scheme improves the overall QoE of clients and QoE fairness among clients based on a state-of-the-art reinforcement learning algorithm. Edge computing assistance plays a role in providing client-side observations to the mobile edge, making agents utilize this information when generating a policy for multi-client adaptive streaming. We evaluated the proposed scheme through simulation-based experiments under various network conditions. The experimental results show that the proposed scheme achieves better performance than the existing schemes.

Various video streaming platforms are responsible for continuously generating a majority of Internet traffic. The promising solution for the smooth video streaming services is to use the HTTP adaptive streaming (HAS). To guarantee a high quality of experience (QoE), the clients using the HAS dedicate the bitrate adaptation for the upcoming video segments. However, the client-driven approach causes degradation in QoE, inefficient resource utilization, and unfair bitrate allocation when multiple clients stream a video over the same access network. Edge computing-assisted adaptive streaming gives an opportunity to jointly optimize the QoE of clients, resource utilization, and fairness among clients by shifting the adaptation intelligence from the clients to the edge cloud. In this paper, we first present an adaptive streaming framework taking advantage of the capabilities of multi-access edge computing (MEC). Next, we design an optimization model that jointly considers the main influencing factors in QoE and fairness among clients. The proposed scheme formulates the joint optimization problem with the various constraints by considering the correlation between QoE and fairness. To efficiently solve the joint optimization problem, we propose a greedy-based bitrate allocation algorithm for multiple clients. The results from the performance evaluation show that the proposed scheme can improve the QoE of clients and resource utilization compared with the existing schemes and minimize the loss in fairness.

Recently, HyperText Transfer Protocol (HTTP) based adaptive streaming (HAS) has been proposed as a solution for efficient use of network resources. HAS performs rate adaptation that adjusts the video quality according to the network conditions. The conventional approaches for rate adaptation involve accurately estimating the available bandwidth or exploiting the playback buffer in HAS clients rather than estimating the network bandwidth. In this paper, we present a playback buffer model for rate adaptation and propose a new buffer-based rate adaptation scheme. First, we model the playback buffer as a queueing system that stores video segments. The proposed scheme selects the next video bitrate that minimizes the difference between the current buffer occupancy and the expected value from the playback buffer model. The evaluation results indicated that the proposed scheme achieves higher video quality than conventional algorithms and can cope with various environments without the tuning of the configuration parameters.