Adaptive sliding window-based video downloading for 1-to-k cooperative SVC streaming using the multi-access edge computing (MEC) architecture

This work proposed the 1-to-k cooperative SVC adaptive streaming method for a group of users based on the Multi-access Edge Computing (MEC) architecture. In the proposed method, a group member’s handheld device is selected as the Header Handheld Device (H-HD), which is in charge of downloading the video content from the MEC Server and then multicasting the downloaded video content to other group members’ Receiver Handheld Devices (R-HDs). The video bit rate that can be adopted in the next transmission cycle is derived using the proposed adaptive Forward Error Correction (FEC) scheme with the estimated bandwidth. To have the smooth streaming, the Currently Downloading Segment Section (CDSS) mechanism was proposed such that the segments inside CDSS can be downloaded in a downloading cycle, i.e., in a time period. CDSS is a sliding window, which can be moved forwardly and whose size can be dynamically adjusted depending on the networking situation, the video playout situation and the buffering situation. A credit scheme was proposed to select the one that has the minimum credit to become the H-HD in next downloading round to balance all handheld devices’ power consumption. Based on the performance evaluation results, the proposed SVC streaming method can have (i) 28.3% higher video quality, which is in terms of the average number of playout video layers and (ii) 52.16% improvement of Quality Of Experience (QOE), which is in terms of the frequency of quality switching, comparing with the target SVC video streaming method; the proposed credit-based H-HD’s re-selection control scheme can have the more even power consumption among all group members’ handheld devices and thus can extend the group’s streaming service time.