SNR aware heterogeneous blind interference alignment in MISO broadcasting channel

The standard Blind Interference Alignment (BIA) accumulates noise levels and causes great Signal Noise Ratio (SNR) deteriorations in the condition that the received SNR is not infinite. To address such problem, we take SNR feedback into consideration. We first define SNR reduced factor, reflecting SNR reduction between the received SNR and the final SNR for decoding data streams. A heterogeneous BIA (hBIA) is then presented where the users form different groups, and each group further consists of several sub-groups. Within one group, the users in one sub-group perform BIA with other sub-groups' users, while the users in different groups have different SNR reduced factors. To maximize the overall effective degree of freedom (DoF) with SNR guarantee, a greedy algorithm is proposed for allocating the users into different groups and subgroups. Evaluations demonstrate that hBIA achieves about 375% and 540% effective DoF improvement on average compared with the standard BIA and the grouping based BIA, respectively.