Pseudo-Lattice Treatment for Subspace Aligned Interference Signals

Abstract: For multi-input multi-output (MIMO) K-user interference networks, we propose the use of a channel transformation technique for joint detection of the useful and interference signals in an interference alignment scenario. We coin our detection technique as "pseudo-lattice treatment" and show that applying our technique, we can alleviate limitations facing Lattice Interference Alignment (L-IA). We show that for a 3-user interference network, two of the users can have their interference aligned in lattice structure through precoding. For the remaining user, performance gains in decoding subspace interference aligned signals at the receiver are achieved using our channel transformation technique. Our "pseudo-lattice" technique can also be applied at all users in case of Subspace Interference Alignment (S-IA). We investigate different solutions for applying channel transformation at the third receiver and evaluate performance for these techniques. Simulations are conducted to show the performance gain in using our pseudo-lattice method over other decoding techniques using different modulation schemes.