Interference Alignment For Mimo Cognitive Networks: A Complex Fdpm-Based Subspace Tracking Approach

Interference alignment (IA) for multiple-input multiple-output (MIMO) cognitive networks is considered by modelling the unlicensed secondary transmitter-receiver pairs which coexist with the licensed multi-antenna primary user as a K-user MIMO interference channel. Starting from investigating the constraint conditions of IA scheme in MIMO cognitive networks, a practical IA algorithm is developed by using a subspace tracking approach based on the complex fast data projection method (FDPM). In the proposed algorithm, first, each secondary transmitter aligns its transmitted signal into the null space of the channel matrix from itself to the primary user without causing any interference to the primary user. Then secondary transmitters and receivers, requiring no channel knowledge of secondary networks, alternately design the precoding and postprocessing matrices through a training period which exploits the complex FDPM-based minor subspace tracking, thus eliminating interference among secondary users. Moreover, the case where secondary transmitters have partial knowledge of channels from themselves to the primary user is also discussed. Simulation results reveal that the proposed algorithm can achieve a high sum rate performance while requiring low computational complexity.