Interference alignment and power control for wireless interference networks

Sammanfattning: This thesis deals with the design of efficient transmission schemes forwireless interference networks, when certain channel state information(CSI) is available at the terminals.In wireless interference networks multiple source-destination pairsshare the same transmission medium for the communications. The signalreception at each destination is affected by the interference from unintendedsources. This may lead to a competitive situation that each sourcetries to compensate the negative effect of interference at its desired destinationby increasing its transmission power, while it in fact increasesthe interference to the other destinations. Ignoring this dependency maycause a significant waste of available radio resource. Since the transmissiondesign for each user is interrelated to the other users’ strategies, anefficient radio resource allocation should be jointly performed consideringall the source-destination pairs. This may require a certain amount ofCSI to be exchanged, e.g. through feedback channels, among differentterminals. In this thesis, we investigate such joint transmission designand resource allocation in wireless interference networks.We first consider the smallest interference network with two sourcedestinationpairs. Each source intends to communicate with its dedicateddestination with a fixed transmission rate. All terminals have the perfectglobal CSI. The power control seeks feasible solutions that properly assigntransmission power to each source in order to guarantee the successfulcommunications of both source-destination pairs. To avoid interference,the transmissions of the two sources can be orthogonalized. They canalso be activated non-orthogonally. In this case, each destination maydirectly decode its desired signals by treating the interference signals asnoise. It may also perform decoding of its desired signals after decodingand subtracting the interference signals sent from the unintendedsources. The non-orthogonal transmission can more efficiently utilize the available channel such that the power control problem has solutions withsmaller transmission power in comparison with the orthogonal transmission.However, due to the randomness of fading effects, feasible powercontrol solutions may not always exist. We quantify the probability thatthe power control problem has feasible solutions, under a Rayleigh fadingenvironment. A hybrid transmission strategy that combines the orthogonaland non-orthogonal transmissions is then employed to use the smallesttransmission power to guarantee the communications in the consideredtwo-user interference network.The network model is further extended to the general K-user interferencenetwork, which is far more complicated than the two-user case. Thecommunication is conducted in a time-varying fading environment. Thefeedback channel’s capacity is limited so that each terminal can obtainonly quantized global CSI. Conventional interference management techniquestend to orthogonalize the transmissions of the sources. However,we permit them to transmit non-orthogonally and apply an interferencealignment scheme to tackle inter-user interference. Ideally, the interferencealignment concept coordinates the transmissions of the sources insuch a way that at each destination the interference signals from differentunintended sources are aligned together in the same sub-space which isdistinguishable from the sub-space for its desired signals. Hence, eachdestination can cancel the interference signals before performing decoding.Nevertheless, due to the imperfect channel knowledge, the interferencecannot be completely eliminated and thus causes difficulties to theinformation recovery process. We study efficient resource allocation intwo different classes of systems. In the first class, each source desires tosend information to its destination with a fixed data rate. The powercontrol problem tends to find the smallest transmission powers to guaranteesuccessful communications between all the source-destination pairs.In another class of systems where the transmission power of each sourceis fixed, a rate adaptation problem seeks the maximum sum throughputthat the network can support. In both cases, the combination of interferencealignment and efficient resource allocation provides substantialperformance enhancement over the conventional orthogonal transmissionscheme.When the fading environment is time-invariant, interference alignmentcan still be realized if each terminal is equipped with multiple antennas.With perfect global CSI at all terminals, the interference signalscan be aligned in the spatial dimension. If each terminal has only localCSI, which refers to the knowledge of channels directly related to the terminal itself, an iterative algorithm can be applied to calculate thenecessary transmitter-side beamformers and receiver-side filters to properlyalign and cancel interference, respectively. Again, due to the lack ofperfect global CSI, it is difficult to completely eliminate the interferenceat each destination. We study the power control problem in this caseto calculate the minimum required power that guarantees each source tosuccessfully communicate with its destination with a fixed transmissionrate. In particular, since only local CSI is available at each terminal, wepropose an iterative algorithm that solves the joint power control andinterference alignment design in a distributed fashion. Our results showthat a substantial performance gain in terms of required transmissionpower over the orthogonalizing the transmissions of different sources canbe obtained.