Abstract
Abstract Carrier Aggregation (CA) is proposed in LTE-Advanced (LTE-A) to allow user equipments (UEs) and LTE-A base station (evolved NodeB, eNB) communicate through multiple carriers simultaneously. Owing to this technique, services utilizing high bandwidth such as high-resolution video streaming become possible in these days. We study a CA-enabled LTE-A video transmission system providing on-demand video service to multiple UEs and formulate a CA-aware radio resource allocation (RRA) problem using buffer length as the performance metric. Video’s buffer length, which is a key factor of the algorithm, is mapped into the Quality-of-Experience (QoE) indicator. We then propose a QoE-based and buffer-aware downlink scheduling algorithm in LTE-A network. It assigns radio resources efficiently and guarantees a lower bound of performance on average QoE. For the scenario that video streaming requests from UE arrive stochastically, we propose a QoE-based access control mechanism to maintain the service quality while balancing the tradeoff between the waiting time and resource efficiency. Our simulation results show significant improvements in QoE of our algorithm over traditional ones.