Multiple-objective cross-layer optimized content-adaptive scheduling for video streaming over 1xEV-DO

In video transmission over wireless cellular packet networks, service fairness, video quality and channel throughput should all be simultaneously guaranteed. In this regard, a key role is played by the scheduling algorithm, that - to achieve maximum performance should consider both physical and application layer information. At the application layer, we propose to consider the video perceptual importance, which depends on both the time-varying semantic relevance and the individual importance of each packet. In this paper, a novel multiple objective optimized (MOO) opportunistic multiple access scheme for slot assignment in a 1xEV-DO system is presented, and modifications to H.264 codec are also described. Here, the user that experiences the best compromise between the least buffer occupancy level, the best channel condition and the highest packet importance indicator is served at each time slot. Results show that this system outperforms the state-of-the-art techniques.