Integer-forcing interference alignment: Iterative optimization via aligned lattice reduction

For the K-user MIMO interference channel, some form of interference alignment is needed to attain the highest possible rates. For the important special case of linear alignment strategies, many iterative optimization algorithms have been proposed that aim to maximize the signal-to-interference-and-noise ratio (SINR) at each receiver. Recent work has demonstrated the advantages of integer-forcing interference alignment (IFIA), which first decodes aligned integer-linear combinations of the codewords and then solves for the desired messages. This paper proposes a class of iterative optimization algorithms for IFIA and demonstrates its advantages via simulations.