Research On Interference Alignment In Wireless Communication Systems

To meet the rising demands of better coverage, higher throughput and lower transmission power, future wireless network is facing a dilemma of limited spectrum resources as well as complex propagation environment. Therefore, new network architectures and transmission mechanisms, such as self-organization network, heterogeneous network, cooperative relay transmission, have been presented. Actually, the introduction of these new network architectures and transmission mechanisms into modern wireless communication systems may incur inevitably severe interference problems. As an innovative wireless transmission strategy that reduces the impact of interference, interference alignment has recently produced many theoretical breakthroughs with respect to analyzing the network capacity. Since the transition of interference alignment from theory to practice still confronts some pressing challenges, the implementation of interference alignment becomes a hot research topic. This dissertation mainly focuses on two practical implementation issues of interference alignment, which are channel state information acquisition and finite number of spatial dimensions for alignment. To deal with the above issues, novel interference alignment schemes are studied and designed. Main work and contributions of this dissertation are summarized as follows:1. Feedback topology designs in multiple-input multiple-output (MIMO) interference channel. Several conventional feedback topologies have been proposed to reduce the channel state information (CSI) acquisition overhead for the implementation of interference alignment in K-user MIMO interference channel (IC). However, these conventional feedback topologies exhibit explicit drawbacks. For the traditional centralized topologies, the computation and feedback burden for all precoders are centralized at either a designated receiver or an extra hardware named as CSI control station. Besides, traditional CSI-exchange topology may incur excessive delay of feedback. In order to reduce the CSI overhead or shorten the time delay for the existing feedback topologies, we propose three new ones, for which the basic ideas are transmitter-receiver grouping, optimizing the sequence of exchanged precoders, and designing brand-new closed-form solution, respectively. Our proposed feedback topologies can not only overcome the drawbacks of the existing topologies, but also achieve a reduction of CSI overhead.2. Antenna resources assignment for MIMO interfering broadcast channel based on interference alignment. By aligning inter-cell interference, interference alignment techniques can greatly improve spatial multiplexing benefits for MIMO interfering broadcast channel. To obtain the same degrees of freedom (DoF), existing closed-form solutions based on interference alignment employ different antenna usages on the base station (BS) side and mobile station (MS) side. For practical concerns, it is unrealistic to install a lot of antennas on a MS, but realizable to deploy a large number of antennas on a massive MIMO BS. Therefore, by flexibly balancing the antenna usage between the BS side and MS side, we propose an interference alignment based antenna resources assignment scheme, which can achieve a tradeoff between the usage of antennas and the DoF gain over orthogonal schemes. Besides, for the cases of2or3users per cell, we provide low-complexity algorithms to align interference, which can significantly reduce the computation complexity while maintaining the same DoF compared to the existing algorithms.3. Antenna usage analysis for interference alignment schemes and closed-form solutions requiring fewer spatial dimensions in heterogeneous network (HetNet). Considering a downlink HetNet, where C picocells coexist in the same spectrum as a macrocell, C pico users and K macro users simultaneously receive the desired signals from their corresponding BSs (d data streams per user), resulting in severe cross-tier interference as well as co-tier interference. For a general framework of the HetNet, the general feasibility conditions of linear interference alignment for closed subscriber group (CSG) and open subscriber group (OSG) modes are investigated, respectively. From the feasibility results, the minimum overall antenna usages for all interference alignment solutions to achieve the DoF of (C+K)d in these two modes are derived, by which we verify that the conventional hierarchical interference alignment (HIA) scheme for CSG mode is optimal in terms of antenna usage. Additionally, it is shown that the conventional HIA scheme for OSG mode is not the optimal antenna utilizing solution, regardless of its antenna configuration and number of data streams per user. Focusing on the general MIMO HetNet in CSG mode, deriving closed-form solutions when restricted to the minimum overall antenna usage is an open problem. Therefore, we relax the minimum antenna utilization constraint to provide a new closed-form interference alignment solution, which outperforms the existing scheme with respect to antenna usage.4. Degrees of freedom for the two-cell two-hop MIMO interference channel. We study the inner bound on the DoF for the two-cell two-hop MIMO interference channel, and propose a new relaying protocol with no global CSI at the source and relay. Achieving a normalized DoF of2/3, the proposed relaying protocol can substantially reduce the decoding complexity at the destination nodes compared to the conventional relaying protocols. Furthermore, we show that the proposed protocol achieves the normalized DoF outer bound of1asymptotically without needing global channel knowledge at the source and relay.