Exploiting Interference Alignment in Multi-Cell Cooperative OFDMA Resource Allocation

Abstract: This paper studies interference alignment (IA) based multi-cell cooperative resource allocation for the downlink OFDMA with universal frequency reuse. Unlike the traditional scheme that treats subcarriers as separate dimensions for resource allocation, the IA technique is utilized to enable frequency-domain precoding over parallel subcarriers. In this paper, the joint optimization of frequency-domain precoding via IA, subcarrier user selection and power allocation is investigated for a cooperative three-cell OFDMA system to maximize the downlink throughput. Numerical results for a simplified symmetric channel setup reveal that the IA-based scheme achieves notable throughput gains over the traditional scheme only when the inter-cell interference link has a comparable strength as the direct link, and the receiver SNR is sufficiently large. Motivated by this observation, a practical hybrid scheme is proposed for cellular systems with heterogenous channel conditions, where the total spectrum is divided into two subbands, over which the IAbased scheme and the traditional scheme are applied for resource allocation to users located in the cell-intersection region and cellnon- intersection region, respectively. It is shown that this hybrid resource allocation scheme flexibly exploits the downlink IA gains for OFDMA-based cellular systems.