Modeling and Analysis of Broadcast-Cellular Converged Networks for Hybrid Live and On-Demand Services

The broadcast-cellular converged network presents a promising architecture for meeting the increasing demand for high-bandwidth and low-latency services while alleviating traffic burdens. In this paper, we consider a broadcast-cellular converged network to provide both live streaming and on-demand services. In this converged network, the broadcast network is responsible for delivering live streaming videos to all users within its coverage area, while the cellular network transmits on-demand videos to intended users. In cases where users cannot successfully receive broadcasting signals, the cellular network also provides live streaming services through network-level traffic offloading. We first analyze and model the live streaming and on-demand video services based on a real-world dataset. The analytical expressions for the average waiting time, system capacity, and power consumption of the proposed broadcast-cellular converged network are then derived based on the $M/G/N/\infty $ queue model. The simulation parameters are based on the actual station deployment scenario in Beijing. We discover that the broadcast-cellular converged network enhances system capacity, reduces average waiting time, and saves energy when the number of users requiring live streaming videos exceeds a certain threshold. We also derive the broadcast on-off threshold, which is independent of the density of on-demand users, but rather dependent on the density of live streaming users. Extensive simulation results demonstrate that compared to the cellular-only network, the proposed converged network can effectively support a larger number of on-demand users and live streaming users simultaneously.