Hybrid blind interference alignment in homogeneous cellular networks with limited coherence time

Abstract

The work presented in this paper is inspired by the user grouping approach of topological blind interference alignment (top-BIA), which is a semi-blind IA scheme and the fact that the partially connected networks can be advantageous in terms of degree of freedom (DoF) and sum rate. The hybrid-BIA scheme proposed in the paper uses top-BIA to group users that are randomly distributed in a dense small-cell network and aims to reduce the supersymbol length and overcome the DoF loss of the state-of-the-art hierarchical BIA (h-BIA) technique. The proposed scheme is very suitable for channels where coherence time is limited and also could attain a good DoF over a small number of symbol extensions. Both the sum-DoF and the network throughput for hybrid-BIA is greater than that of h-BIA. By varying the number of user groups, the paper shows that h-BIA constitutes a special case of hybrid-BIA. Finally, the paper demonstrates the effect of changing the number of transmit antennas and number of small cells on the sum-DoF gain of hybrid-BIA over that of h-BIA.