Data-Driven Adaptive Network Slicing for Multi-Tenant Networks

Network slicing to support multi-tenancy plays a key role in improving the performance of 5G networks. In this paper, we propose two time-scale framework for reservation-based network backhaul and Radio Access (RAN). proposed scheme, subset slices is activated via novel sparse optimization long with goal maximizing expected utilities tenants while short are reconfigured according time-varying user traffic channel states. Specifically, using statistics from users channels also considering utility serving slice reconfiguration cost, formulate problem update configuration resources such that maximum isolation reserved enforced. The formulated problems time-scales non-convex difficult solve. We use $\ell_q$-norm, $0<q<1$, group LASSO regularizations iteratively find convex approximations problems. Frank-Wolfe algorithm solve approximated time-scales. To cope dynamical nature variations, fast, distributed Simulation results demonstrate our approaches relative optimal solutions existing state art method.