Optimal power allocation for secondary users in CR networks

Abstract

Cognitive Radio (CR) increases spectrum utilization by allowing unlicensed secondary users to concurrently operate with the licensed primary users. The side-lobes of an OFDM-based secondary user (SU) causes unwanted interference to the sub-carriers of OFDM-based primary user (PU). The objective is to assign power to the sub-carriers of SU in such a way that capacity of SU is maximized while satisfying the total power budget constraint and interference (introduced to the PU from SU) constraint. Previously, this power allocation to the SU's sub-carriers (i.e. convex optimization algorithm) was solved using KKT conditions & iterative algorithms. These algorithms maximize capacity asymptotically when the iterations are very large. In this paper, closed-form solution is presented that calculate the power allocation to the SU's sub-carriers. Simulation results are included which shows that complexity of the closed-form solution is small when compared to the complexity of the iterative algorithms.