Heterogeneous Semi-Blind Interference Alignment in Finite-SNR Networks With Fairness Consideration

摘要

Standard blind interference alignment (sBIA) suffers from noise accumulation which severely deteriorates received signal-to-noise ratio (SNR) and significantly reduces transmission rate. A noise accumulation factor is proposed to describe the loss between the user & x2019;s received SNR, and the final post processing SNR which determines the performance of the decoding of the encoded data streams (EDSs). A heterogeneous semi-BIA (H-SBIA) framework where users with different noise accumulation factors can be flexibly allocated effective EDSs (E-EDSs) is constructed. Relying on the H-SBIA framework, a heuristic H-SBIA algorithm is designed for maximizing the overall E-EDSs considering both fairness and coherence time constraints. Extensive simulations demonstrate that H-SBIA produces great fairness performance improvement at a limited cost in the achievable sum rate. The Jain & x2019;s fairness index is about 2.2 times greater than that for SNR-SBIA proposed in previous work, at the cost of sacrificing 10 & x0025; of the achievable sum rate.