Joint Beamforming and Power Allocation for Throughput Maximization in Cognitive Radio

Abstract

We consider the total throughput maximization problem involving joint beamfoming and power allocation in the cognitive radio network (CN), wherein the single-input multi-output multiple access channels are assumed. Subject to two sets of constraints: the interference constraint of primary user (PU) and the peak power constraints of cognitive users (CUs), a joint beamforming and power allocation algorithm for total throughput maximization in CN is proposed, which is implemented by alternating iteration of MMSE-DFE beamforming and a best response based iterative power allocation algorithm, we note it as MMSE-BR. Compared with the existing algorithm, the proposed MMSE-BR can achieve the optimal throughput under various interference situation, and does not have limit on the number of CU. The optimality of MMSE-BR algorithm is detailedly proved in theorem and corroborating numerical results are presented to present the comparisons with the existing algorithm.

Joint Beamforming and Power Allocation for Multiple Primary Users and Secondary Users in Cognitive MIMO Systems via Game Theory

“…The alternating iteration is a low-complexity method for multi-variable optimization problem and used widely. We employ an iterative algorithm that repeats two sets of optimization variables: beamforming matrix w and transmission power vector p until convergence to solve the optimization problem [20].…”

“…Inspired by these considerations, in this paper, under a game-theoretic framework, we study the problem of joint beamforming and power allocation in a cognitive MIMO system wherein multiple primary users and secondary users are co-located. We design an algorithm based on the minimum mean square error based decision feedback (MMSE-DFE) beamforming algorithm and a best response based (BR) iterative power control algorithm [20]. The proposed algorithm can achieve better throughput performance considering the interference and noise power, and does not limit the number of active users.…”

“…condition(20) of SP2 cannot be satisfied, so (0) p  is the optimal solution to(16). If (0) p  destroys the interference constraints of PUs, it is necessary to solve SP2, and the solution to Copyright ⓒ 2013 KSII…”