Joint resource allocation for learning-based cognitive radio networks with MIMO-OFDM relay-aided transmissions

In this paper, we investigate the joint power allocation for dual-hop amplify-and-forward (AF) MIMO-OFDM cognitive radio (CR) networks. The considered AF MIMO-OFDM CR network coexists with a primary radio (PR) network through underlay spectrum sharing. In order to mitigate the interference to the PR network, environmental learning algorithm is adopted to blindly estimate the null space of the PR user, which are orthogonal to the PR communication channels. With necessary channel state information, under independent transmit power constraints as well as the interference constraints, the power allocation of CR source and relay and subcarrier pairing over two hops are optimized jointly to maximize the CR network throughput. Furthermore, the relay node implements an effective subcarrier permutation policy to enhance the performance further at the cost of affordable complexity. Finally, the performance advantages of the proposed algorithm are demonstrated by the simulation results.