Relaxing the CSIT Requirement for Alignment-based Interference Management in Multiuser Wireless Communication Systems

As the demand for higher data rate wireless communications keeps on increasing in the past decades, there is a need for more efficient utilization of the scarce spectral resource in the system. Several methods including universal frequency reuse or deploying smaller cells have been proposed. However, these methods also induce strong interference in the multiuser communication scenario, which, if not being managed properly, may degrade the system performance dramatically. Hence, how to manage interference has been a major challenge in the wireless system design.;Recently, alignment-based approaches have been proposed as a novel strategy to reduce the impact of interference. The idea is to cooperatively design the transceiver strategies so that interference-free signal spaces can be created by aligning interference or signal. One potential issue which needs to be addressed before applying alignment-based approaches in practical scenarios is that the cooperative transceiver design for alignment requires extensive channel state information (CSI), which could be difficult to acquire, especially at the transmitters. As an attempt to solve this problem, in this thesis, we provide some interesting ideas on relaxing the CSIT requirement for alignment-based interference management in multiuser wireless communication systems.;Firstly, we focus on the achievable degrees of freedom (DoF) of a fully connected multiple-input multiple-output (MIMO) interference channel with a MIMO relay. We find that relay eliminates the CSIT requirement at the transmitters and helps to achieve the optimal DoF through relay-aided interference alignment (IA). Interestingly, the processing at the relay only involves linear processing which represents a major reduction in the complexity as compared to other schemes which require decoding at the relay.;We also investigate the application of spatial IA in a multiuser interference network with asymmetric channel attenuation at finite signal-to-noise ratio (SNR). We exploit the asymmetry in the path-loss and propose to use clustering as a scheduling mechanism to address the issue of IA feasibility constraint and signaling overhead. In particular, we show that proper clustering combined with generalized precoder/equalizer design leads to effective suppression of interference. Moreover, the signaling overhead is relatively low since the acquisition of the instantaneous CSI for IA is only limited within separated clusters.;Finally, we present the idea of utilizing relay selection to alleviate the CSIT requirement for signal space alignment (SSA) based interference management in a multiuser cellular two-way relay system. We show that exploiting the opportunities of having multiple relays eliminates the CSIT requirement to implement SSA. We propose a relay selection scheme which only requires single real-valued report from each relay, thus it has the advantage of low signaling overhead. A sufficient condition for the number of relays to scale with the transmission power to maintain the optimal DoF is also provided.