A novel channel predictor for interference alignment in cognitive radio network

In a cognitive radio (CR) network, how to eliminate the interference between primary users and secondary users is a curial work. The emergence of interference alignment (IA) provides an effective way to solve this problem. However, in order to utilize the IA algorithm, the real-time and accurate channel state information (CSI) is required at both transmitters and receivers. But in practical IA system, it is hard to get the perfect CSI due to the capacity constraint, channel estimation errors and time delay, which will severely affect the system performance. In this paper, the impact of delayed CSI on average signal to interference plus noise ratio (SINR) and achievable sum rate of IA system are analyzed. To eliminate the effect of delayed CSI, a novel channel predictor based on the linear Markov chain (LMC) is proposed. Using the finite state Markov chain model, the CSI of next time instant can be predicted according to the former and current CSI. Simulation results show that the proposed IA scheme based on LMC predictor can significantly upgrade the performance of IA system with the delayed CSI, and it can achieve better results with lower complexity compared with traditional AR predictor.