Efficient Resource Allocation for MIMO-OFDM Cognitive Networks with Adaptive Precoding

Multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) combined with cognitive radio is a promising technology for future networks. In this paper, we study the resource allocation in MIMO-OFDM based cognitive radio cellular networks with capacity-achieving adaptive precoding (capacity-achieving in the sense that it approaches the capacity upper-bound of the CR link when the primary radio link is in idle state). Since the formulated optimization task defines a mixed integer programming problem of exponential complexity, we propose a two-phase method to produce an optimal yet efficient solution. Particularly, by jointly considering the signal-tonoise ratios (SNRs) of MIMO-OFDM subcarriers and the interference power introduced to primary users (PUs), we propose an efficient subcarrier allocation procedure and an adaptive precoding selection strategy. Furthermore, we develop a rapid, stable and efficient algorithm to optimally distribute the power among the subcarriers using convex optimization. Numerical results show that our proposal can outperform other standard techniques with respect to the sum-rate and the reduction in complexity.