The evolution of wireless communication must meet the increasingly high demand in mobile data. It is expected to increase the maximum rates of wireless by a factor of 1000 by 2020. Meanwhile, it is clear that to reach this goal, a combination of different ingredients is necessary. The major limitation of wireless systems is the interference due to frequency reuse. This has been a long-standing impairment in cellular networks of all generations that will be further exacerbated in 5G networks, due to the expected dense cell deployment.
The use of orthogonal frequency-division multiplexing (OFDM) in 4G leaded to an interference management by dynamic coordination of resource blocks. However, this allowed only modest gains in rates. A new technique of interference management was born 5 years ago, the Interference Alignment (IA). The IA permits to have a capacity with equals the half of the capacity of an interference-free system. This technique supposes that each transmitter (Tx) knows the channels not only towards its receivers (Rx)s, but the channels from all Txs to all receivers Rxs.
A more recent interference technique that boosts IA is Massive Multiple Input Multiple Output (MIMO), where Txs use antennas at a very large scale. The idea is motivated by many simplifications, which appear in an asymptotic regime where base stations are endowed with large numbers of antennas.
This thesis treats the problem of interference cancellation and capacity maximization in Massive MIMO. In this context, the thesis proposes new interference management alternatives for the Massive MIMO antenna regime, taking into account also the practical challenges of Massive antenna arrays and focusing mainly on the lack of perfect channel knowledge at the Tx side.
