In this paper, we address dynamic spectrum access for Primary Operators (POs) for LTE-Advanced systems. We propose a dynamic spectrum access framework that exploits the capabilities of carrier aggregation to efficiently utilize the unutilized spectrum which varies with time and space when static spectrum access policies are adopted. A spectrum owner (SO) adopts an auction scheme for accessing spectrum for the dynamic requests. Furthermore, we introduce an accurate model driven by real-life statics and LTE-A standard specifications for estimating the bandwidth required to satisfy the traffic demands of an operator’s subscribers for a projected demand model.

For the auction process, we map the spectrum access problem into a bounded knapsack problem which is solved using dynamic programming in pseudo-polynomial time. The solution uses the generalized second price strategy. Simulation results show more than 20% reduction in the required spectrum and up to 80% reduction in the average unutilized spectrum as compared with traditional static spectrum access due to the on-demand spectrum assignment.