Performance Enhancement of Multi User Multi Cell Interference Alignment with Pair Selection

Spectral efficiency of a multi user MIMO system can be enhanced using user selection diversity. In this work we exploit the selection diversity to improve the performance of multi user (MU) interference alignment (IA) based transmit precoding in downlink MIMO multi cell system. We propose three pair selection algorithms which can be categorized based on their feedback requirements and computational complexity. Our algorithms are suboptimal but we show with the help of extensive simulations that they provide considerable gains in the performance of IA based precoding. The algorithms are also compared with a standard pair selection algorithm. The results show that our algorithms outperform the standard selection. We also assess the performance of IA based precoding with user selection by comparing it with baseline precoding in two cellular scenarios. The two scenarios differ with respect to the inter cell interference characteristics. The comparison shows that in both cases IA based precoding outperforms the baseline using our proposed selection algorithms.