A Context-aware Radio Resource Management in Heterogeneous Virtual RANs

New-generation wireless networks are designed to support a wide range of services with diverse key performance indicators (KPIs) requirements. A fundamental component of such networks, and a pivotal factor to the fulfillment of the services target KPIs, is the virtual radio access network (vRAN), which allows high flexibility on the control of the radio link. However, to fully exploit the potentiality of vRANs in non- stationary environments, an efficient mapping of the rapidly varying context to radio control decisions is not only essential, but also challenging owing to the interdependence of user traffic demand, channel conditions, and resource allocation. In this paper, we propose CAREM, a reinforcement learning framework for dynamic radio resource allocation in heterogeneous vRANs, which selects the best available link and transmission parameters for packet transfer, so as to meet the KPI requirements. To show its effectiveness in real-world conditions, we provide a proof-of- concept through a testbed implementation. Experimental results demonstrate that CAREM enables an efficient radio resource allocation, for different time periodicity of the decision-making process as well as under different settings and traffic demand. Furthermore, the results show that CAREM outperforms state- of-the-art solutions as well as standard cellular technologies: when compared to the closest existing scheme based on neural network and the standard LTE, it exhibits an improvement of about one order of magnitude in both packet loss and latency, while it provides a 65% latency improvement with respect to relatively simpler and more popular contextual bandit approach.