Resource Allocation for Medium Grain Scalable Videos over Femtocell Cognitive Radio Networks

Femtocells are shown highly effective on improving network coverage and capacity by bringing base stations closer to mobile users. In this paper, we investigate the problem of streaming scalable videos in femtocell cognitive radio (CR) networks. This is a challenging problem due to the stringent QoS requirements of real-time videos and the new dimensions of network dynamics and uncertainties in CR networks. We develop a framework that captures the key design issues and trade-offs with a stochastic programming problem formulation. In the case of a single FBS, we develop an optimum-achieving distributed algorithm, which is shown also optimal for the case of multiple non-interfering FBS's. In the case of interfering FBS's, we develop a greedy algorithm that can compute near-optimal solutions, and prove a closed-form lower bound for its performance. The proposed algorithms are evaluated with simulations, and are shown to outperform two alternative schemes with considerable margins.