Block diagonalization with subspace based external interference mitigation

Abstract

In this work we propose a variation to the Block Diagonalization (BD) algorithm described that mitigates the interference from an external interference source which does not participate in the joint transmission inherent to the BD algorithm. The idea consists of precoding the signal at the transmission side so that the received signal at each receiver lies in a subspace orthogonal to the subspace containing the external interference. It is similar to how interference alignment algorithms work, but here instead of limiting the caused interference at the unintended receivers to a specific subspace, we are limiting the desired signal at the intended receivers. Since this “limiting” aspect reduces the dimensions available for the useful signal, a trade-off is involved between how many useful streams can be transmitted and how much of the external interference energy can be canceled. The proposed algorithm is analyzed and we show different metrics that can be used to determine the “good point” in this trade-off.