Resource allocations for MISO based cognitive radio OFDMA systems

We propose an optimal and a suboptimal resource allocation algorithms for sum rate maximization in an MISO-OFDMA based cognitive radio network subject to a primary user outage constraint. By converting the outage constraint into user and subcarrier dependent peak power constraints, we obtain the optimal joint subcarrier assignment and power allocation solution by applying Lagrange dual-decomposition method. We also propose a low-complexity, fast-convergent suboptimal algorithm which uses an opportunistic channel gain-to-noise ratio based subcarrier assignment and an efficient constrained power allocation scheme. The optimality of the latter power allocation scheme is verified as well. Numerical results demonstrate that, compared with the optimal solution, the suboptimal scheme suffers only minor performance loss. We also shown that the use of more transmit antennas by a cognitive base station not only yields better spectrum efficiency but also lessens the outage-induced peak power constraint.