Adaptive Bandwidth Sharing for Optimizing QoE of Real-Time Video

Abstract: The concept of spectrum or bandwidth sharing has gained significant global attention as a means to enhance the efficiency of real-time traffic management in wireless networks. Effective bandwidth sharing enables optimal utilization of available resources, reducing congestion and improving QoE for delay-sensitive applications such as real-time video transmission. In this paper, we propose a novel iterative semi-static bandwidth sharing policy that balances the advantages of both static and dynamic sharing approaches. Our approach minimizes the frequency of coordination between network operators while ensuring efficient resource allocation and meeting the stringent QoE demands of real-time traffic. The proposed policy iteratively optimizes both the spectrum sharing between operators and the resource allocation for individual clients. We establish strong theoretical guarantees for the optimality of the proposed policy and prove that it converges to the optimal static sharing policy irrespective of initial conditions or fluctuations in traffic arrival rates. Additionally, we conduct extensive simulations to evaluate the impact of key system parameters - including step size, hyperperiod length, and arrival process dynamics - on the performance of our policy. Our results demonstrate the effectiveness of the proposed approach in achieving near-optimal bandwidth allocation with reduced overhead, making it a practical solution for real-time wireless applications.