A quaternion-based approach to interference alignment with Alamouti coding

Based on the representation of Alamouti space-time codewords as quaternions, this paper proposes a scheme that combines interference alignment (IA) with Alamouti signals. The proposed formulation allows for a separation of the space-time block coding (to gain spatial diversity) and the IA precoding (to reduce or ideally suppress interference). Although this separation is not necessarily optimal, the splitting of alignment precoding and Alamouti encoding is particularly convenient because it enables the independent optimization of the IA solution using quaternionic versions of standard alternating optimization techniques such as the maximum signal-to-interference-plus-noise algorithm. Some numerical simulations are included to compare the performance of the proposed quaternion IA+Alamouti algorithm with standard IA algorithms in the complex domain as well as with interference cancellation schemes at the receiver side.