On Optimal Orchestration of Virtualized Cellular Networks With Downlink Rate Coverage Probability Constraints

Wireless network virtualization is emerging as a promising technology for next-generation (5G) cellular networks. A key advantage of introducing virtualization in cellular networks is that wireless services can be decoupled from network resources (e.g., infrastructure and spectrum) so that multiple virtual networks can be built using a shared pool of network resources. This paper develops a theoretical framework for optimizing the resource allocation in virtualized cellular networks with heterogeneous coverage requirements. Specifically, we first formulate a chance-constrained virtual resource allocation problem that aims at probabilistically guaranteeing virtual networks' downlink coverage and rate demand satisfaction while minimizing resource over-provisioning in the presence of uncertainty in user equipment (UE) locations and channel conditions. Thereafter, we derive a closed-form expression for the downlink rate coverage probability of a typical virtual network. With the closed-form expression, we design an efficient algorithm to solve the chance-constrained problem with affordable computation complexity. Furthermore, considering the possibility of lack of sufficient network resources to satisfy all virtual networks' demands, we design a prioritized virtual resource allocation scheme where virtual networks are built sequentially based on their given priorities. Our results demonstrate that the proposed stochastic virtualization framework outperforms existing deterministic virtualization frameworks in terms of probabilistically guaranteeing virtual networks' coverage and rate demand satisfaction.