Joint Optimization of User Grouping and Transmitter Connection on Multi-Cell SNR Blind Interference Alignment

Blind interference alignment (BIA) can greatly improve the degree of freedom with the infinite signal-to-noise ratio (SNR) assumption. Under the finite SNR condition, noise accumulation can have a significantly negative impact on SNR, inducing severe performance deterioration. In particular, in multi-cell networks, the transmitter to which a user connects can further affect its received SNR and the BIA design. To address such problem, we present a user grouping scheme for reducing noise accumulation in a single cell and analyze the impact of transmitter connections on the user grouping scheme. SNR BIA in a multi-cell network is further proposed, which jointly optimizes the transmitter connection and the user grouping scheme. Extensive simulations demonstrate that the achievable sum rate of SNR BIA is 1.36 times, 1.66 times, and 2.68 times that of data shared BIA, standard BIA, and extended BIA reported in the literature, respectively, and SNR BIA is more robust to user mobility.