Toward Optimal DoF Maximization With Interference Categorization Using TIM and SIC

In this paper, a novel topological interference management (TIM) framework, combined with the successive interference cancellation (SIC) technique, is proposed for Device-to-Device (D2D) networks. While the TIM problem was originally studied in a partially connected network, the main contribution appears here in proposing a novel design that allows some interfering links to be very high so that SIC can be used to handle them efficiently. More specifically, this inter-twinning between the two techniques (TIM and SIC) is based on the design of the precoding and decoding vectors used by the transmitters and receivers, respectively. This scenario of having very strong interference may occur in crowded D2D networks, such as malls, concerts or stadiums, where D2D communications take place in very close proximity. The problem is formulated as a rank minimization framework and solved by building on the characteristic polynomial function (CPF) of the adjacency matrix of the network interference graph. Numerical results show the effectiveness of our approach by achieving promising gains in terms of degrees-of-freedom (DoF), as compared to existing solutions, with a polynomial complexity.