Introduction - Interference Cancellation Using Space-Time Processing and Precoding Design - page 3

limited feedback is utilized to adapt the phase of a transmitted signal and improve the
performance of the system. However, to the best of our knowledge, there is no result
showing how to achieve full diversity and interference cancellation for each user
using limited feedback. A naive way is to quantize the result in [ 18 ] directly. But this
will not work because the scheme in [ 18 ] relies on the perfect channel information
and thus perfect orthogonality between the signal vectors of the two users. Simply
quantizing the results will destroy the perfect orthogonality and thus cannot achieve
full diversity. In this topic we investigate how to use quantized feedback to achieve
full diversity as well as interference cancellation. Our results show that even with
quantized feedback, full diversity and interference cancellation are possible by using
our proposed scheme. Also our decoding complexity is the lowest to our best knowl-
edge. By increasing the number of feedback bits, the performance of our proposed
scheme will approach the performance of the scheme with perfect feedback in [ 18 ].
So our proposed scheme can serve as a bridge between the schemes with no feedback
and perfect feedback.
1.3 Interference Cancellation and Detection for X Channels
When there are two users each transmitting different codewords to two receivers
simultaneously, a scenario known as X channels, multiple antennas can be utilized
to increase the date rate as discussed in the existing literature. For example, the
schemes proposed in [ 19 , 20 ] can achieve the highest multiplexing gain with no or
partial cooperation between users. References [ 21 , 22 ] provide the achievability as
well as converse results for the degrees of freedom region of a MIMO X channel
using a technique called interference alignment when perfect channel knowledge
is available to all transmitters and receivers. Capacity region and Relay-Aided X
channels are discussed in [ 23 , 24 ].
The main emphasis of the above papers has been the maximummultiplexing gain.
But, in most cases, these schemes achieve a diversity of one. On the other hand, in
a system with limited complexity and delay constraints, reliability, in terms of error
probability or diversity, is also important. Interference cancellation methods have
been proposed to suppress the interference from other users and reduce the error
probability in multiple access channels with limited delay and complexity [ 2 , 4 , 5 ,
9 , 25 ]. Unfortunately, these interference cancellation methods cannot be used in X
channels.
We investigate the following two problems: (1) how to realize interference-free
transmission for each user to reduce the error probability and the decoding com-
plexity. (2) how to achieve the highest possible diversity to improve the transmission
quality in X channels. To the best of our knowledge, the only way to achieve full
diversity for each user in X channels is to jointly decode the codewords from all users
using maximum-likelihood decoding. The decoding complexity of such a scheme
is very high. Also, each receiver will have access to the information of the other
receiver which may not be desirable. Our proposed precoding and decoding scheme