Memory and memoryless optimal time-window controllers for secondary users in vehicular networks

In this paper, a primary-secondary resource-management controller on vehicular networks is designed and tested. We formulate the resource-management problem as a constrained stochastic network utility maximization problem and derive the optimal resource management controller, which dynamically allocates the access time-windows to the secondary-users. We provide the optimal steady-state controllers under hard and soft primary-secondary collision constraints, showing as the hard controller does not present any optimality gap in the average utility respect to the soft one, while, on the contrary, it is able to make the outage-probability vanishing. Then, we present as a particular case the subset of memoryless controller, that are unable to exploit the system statistics, derive the throughput-gain of the general controllers with respect to the memoryless ones and discuss conditions of applicability and advantages of each subclass.