Can interference alignment impact network utility maximization?

This paper examines whether interference alignment (IA) can be leveraged to improve network utility in a multiuser multi-antenna wireless cellular network. Optimality of IA from a DoF standpoint has the potential to aid conventional network optimization algorithms that typically can only find locally optimal solutions. This paper investigates the usefulness of IA for interference coordination and utility maximization by proposing a two-stage optimization framework for a G-cell multi-antenna network with K users/cell, and with full channel state information (CSI) available at all base-stations. The first stage of the proposed framework focuses exclusively on nulling interference from a set of dominant interferers using IA, while the second stage optimizes the transmit and receive beamformers iteratively to maximize a network-wide utility using the IA solution as the initial point. The number of dominant interferers to be nulled in the first stage is guided by a set of new feasibility results for partial IA. This paper focuses on maximizing the specific network utility of minimum rate over all users in the network. Through simulations on two different topologies of cluster of BSs either in isolation or in the presence of other non- cooperating BSs, the proposed framework with IA initialization is observed to outperform straightforward optimization on an isolated cluster of BSs. But, IA loses its impact when there is significant out-of-cluster interference. Thus, in a large-scale dense cellular deployment, the benefit of IA is likely to be limited, even with centralized network optimization and full CSI.