An Efficient Prewhitening Scheme for MIMO Cognitive Radio Systems

In this paper, we investigate an interference mitigation scheme under multiple-input multiple-output (MIMO)-based cognitive radio environment, where primary and secondary systems cooperate, sharing the channel state information (CSI) of the interference link. We propose a prewhitening scheme where the precoder at a secondary transmitter uncorrelates the correlated interference in advance using only the direction information on the interference channel reported from a primary system. First, we decompose the precoding matrix of a conventional prewhitening scheme, in which full CSI is required, into direction and magnitude components. Then, using only the direction information on the interference channel, the proposed precoder projects the row-space vectors of the interference channel onto the scaled null-space vectors from the precoder. Through mathematical analysis and simulations, we show that the proposed prewhitening scheme can achieve almost the same performance as the conventional prewhitening scheme. We investigate system performance in terms of achievable rate, average received interference power, and average signal-to-interference ratio. As a result, we can conclude that for prewhitening, the use of the direction information on the interference channel can provide almost the same performance as that of the full CSI.