Stochastic Precoding for MISO Interference Channels with Channel Mean Feedback

Two linear precoding strategies are investigated for multiple-input single-out (MISO) interference channels with channel mean feedback at transmitters. The interference at each receiver is treated as additive noise. The challenge here is that previous analysis with perfect channel state information (CSI) at transmitters does not apply. Based on stochastic ordering, the altruistic non-equilibrium stochastic zero forcing strategy is first proposed, generalizing traditional zero forcing which requires perfect CSI. Interestingly, the precoding matrices here are all rank-one as in traditional zero forcing. The competitive use of the common channel is also formulated as a strategic game. Different from the perfect CSI case, with channel mean feedback, infinitely many Nash equilibria exist in general, which are not necessarily rank-one. Comparisons of the above two precoding strategies reveal no overall dominance of one over the other, thereby establishing stochastic zero forcing as an appealing alternative to the prevailing Nash equilibrium strategies.